INDEX

---

Page	Section
1	Portada
2	Index
3	Glossary of Metrics
4	Executive Summary
5	Context and Methodology
6	Competitive Benchmark: Share of Voice
7	Competitive Benchmark: Brand Impact Score
8	Competitive Benchmark: Sentiment
9	Brand Confusion
10	Co-occurrence and Ecosystem
11	Perception by Business Area
12	Territorial Bias
13	Analysis by Persona
14	Analysis by Funnel
15	Verbatims
16	Sources Cited by the AI
17	Owned Media: GEO Evaluation
18	Narrative Control Strategy
19	SWOT Analysis
20	GEO Scorecard
21	Strategic Imperatives
22	90-Day Roadmap
23	Conclusion

---

Note: This report analyses 18,200 responses generated by ChatGPT and Gemini using 9,720 prompts created with the scientifically validated Prompt Atlas methodology. Study covers 4 markets (US, Australia, Spain, Middle East), 3 product lines, 3 funnel stages, and 36 personas across 15 defined competitors.

GLOSSARY OF GEO METRICS

---

Core Metrics

Mention Score (MS)

The proportion of brand mentions relative to total mentions detected across all entities in a given response.

• Scale: 0 to 1
• Formula: entityMentionCount / totalMentionCount
• Interpretation: Higher values indicate greater relative presence across all AI responses. A score of 0.23 means the brand accounts for 23% of all entity mentions in responses where it appears.

Veolia (ChatGPT): 3,290 / total mentions = 0.226 Veolia (Gemini): 4,555 / total mentions = 0.154

---

Position Score (PS)

The brand's share of positional importance within each response, weighted with logarithmic decay — earlier mentions carry more weight.

• Scale: 0 to 1
• Formula: entityWeightSum / totalWeightSum, where weight(position) = 1 / log(position + 2)
• Interpretation: Position 0 (first mention) is worth 3.5× more than position 10. Brands mentioned first have greater cognitive impact on the reader.

Position	Weight
0 (first)	1.443
1	0.910
2	0.721
5	0.512
10	0.417

Veolia (ChatGPT): avg mention position 2.26 → Position Score 0.269 Veolia (Gemini): avg mention position 2.70 → Position Score 0.214

---

Sentiment Score (SS)

The tone of brand mentions, normalised from the raw 1–5 detector scale.

• Scale: -1.0 (very negative) to +1.0 (very positive)
• Formula: (sentiment_raw - 3) / 2

Raw Sentiment	Normalised Value
1 (very negative)	-1.0
2 (negative)	-0.5
3 (neutral)	0.0
4 (positive)	+0.5
5 (very positive)	+1.0

Veolia (ChatGPT): +0.381 — consistently positive framing Veolia (Gemini): +0.417 — slightly more positive tone

---

Competitive Score (CS)

Relative visibility ranking among all entities detected in the same responses.

• Scale: 0 to 1
• Formula: (totalEntities - ranking + 1) / totalEntities
• Interpretation: 1.0 = absolute leader, 0.0 = last place. Measures how Veolia performs head-to-head against every other brand the AI mentions in the same response.

Veolia (ChatGPT): 0.777 — Veolia outranks 77.7% of all co-occurring entities Veolia (Gemini): 0.843 — Veolia outranks 84.3% of all co-occurring entities

---

Brand Impact Score (BIS)

A composite executive indicator integrating all four metrics into a single 0–100 score.

• Formula: ``` normalizedSentiment = (sentimentScore + 1) / 2

BIS = (normalizedSentiment × 0.30 + positionScore × 0.25 + mentionScore × 0.25 + competitiveScore × 0.20) × 100 ```

Weight	Component
0.30	Sentiment (normalised to 0–1)
0.25	Position Score
0.25	Mention Score
0.20	Competitive Score

Veolia BIS — ChatGPT (calculated from real values):

normalizedSentiment = (0.381 + 1) / 2 = 0.690
BIS = (0.690 × 0.30 + 0.269 × 0.25 + 0.226 × 0.25 + 0.777 × 0.20) × 100
    = (0.207 + 0.067 + 0.057 + 0.155) × 100
    = 0.486 × 100
    = 48.6 ✓

Veolia BIS — Gemini (calculated from real values):

normalizedSentiment = (0.417 + 1) / 2 = 0.708
BIS = (0.708 × 0.30 + 0.214 × 0.25 + 0.154 × 0.25 + 0.843 × 0.20) × 100
    = (0.212 + 0.054 + 0.039 + 0.169) × 100
    = 0.473 × 100
    = 47.3 ✓

BIS	Meaning
80+	AI recommends you first, spontaneously
60–80	Well positioned, consistently visible
40–60	Present but not dominant
<40	Competitors outpace you

Veolia sits at 48.6 (ChatGPT) and 47.3 (Gemini) — present but not dominant. Clear room to grow.

---

Share of Voice (SOV)

Absolute penetration: the proportion of all AI responses that mention the brand at least once.

• Formula: (Responses featuring brand / Total responses) × 100
• Interpretation: Directly comparable to offline market share data to detect gaps between real market position and AI-perceived position.

Veolia (ChatGPT): 2,066 / 9,374 = 21.3% Veolia (Gemini): 2,736 / 8,826 = 29.0%

---

Share of Branded Voice (SBOV)

Relative market share in responses where at least one brand is mentioned — competitive share within the branded universe.

• Formula: (Brand mentions / Total branded mentions) × 100
• Interpretation: If SBOV < real market share → AI underrepresents the brand relative to competitors.

Veolia (ChatGPT): 22.0% Veolia (Gemini): 31.0%

---

Key GEO Concepts

Prompt Atlas

Scientific prompt generation engine that creates neutral prompts for visibility studies. Generates combinations of persona × funnel × product line using validated methodology. Prompts never mention the analysed brand — they measure whether AI recommends it spontaneously.

GeoRadar

The audit engine that runs prompts against multiple LLMs (ChatGPT, Gemini, Perplexity, Claude) and analyses visibility, tone, and competitive positioning.

S.A.M. (Semantic Alignment Machine)

Semantic content validation system. Given any text, it finds the most similar study prompts using embeddings (inverse RAG). Measures content coverage, detects gaps, and predicts affinity gains before publishing.

---

AI Engines Analysed

• ChatGPT (OpenAI)
• Gemini (Google)

---

Formulas source: references/metrics/entity-metrics-system.md — the single source of truth for GeoRadar metric definitions.

Page 4 — Executive Summary

---

What This Audit Measures

The Prompt Atlas generates 9,720 unique prompts simulating real user queries across Veolia's three business areas: Water Technologies & New Solutions, Hazardous Waste Treatment, and Bioenergy & Energy Efficiency. Approximately 96% of prompts are neutral — they do not mention "Veolia" by name. They cover 4 markets (United States, Australia, Spain, Middle East), 3 funnel stages, and 36 persona profiles.

When a procurement manager in Sydney asks ChatGPT "Which companies offer industrial water treatment for mining operations?" or a municipal engineer in Madrid asks Gemini "¿Quién lidera el tratamiento de residuos peligrosos industriales en Europa?", the AI chooses which companies to mention. This audit measures whether Veolia appears in those responses, in what position, with what tone, and how often — across all contexts where it should be visible.

---

The Numbers

Global Visibility

Metric	ChatGPT (9,374 responses)	Gemini (8,826 responses)
Veolia mentions (total)	3,290	4,555
Responses featuring Veolia	2,066 (21.3%)	2,736 (29.0%)
BIS (Brand Impact Score)	48.6	47.3
SBOV (among brands)	22.0%	31.0%
Sentiment Score	+0.381	+0.417
Avg mention position	2.26	2.70
Competitive Score	0.777	0.843

---

What the Data Reveals

1. Veolia leads all 15 defined competitors by a wide margin — but the gap to dominance is significant

Veolia is the most visible brand in this study among all 16 entities (itself plus 15 competitors) by Share of Voice in both engines. However, the BIS gap between Veolia and category dominance (80+) is still 31 points. The AI mentions Veolia, but rarely as the first and definitive answer.

Engine	Veolia BIS	Competitor Pool Avg BIS	Gap (Veolia vs Pool)
ChatGPT	48.6	41.6	+7.0
Gemini	47.3	37.0	+10.3

Closest competitors in ChatGPT by BIS: ENGIE (49.2), Ameresco (48.4), Clean Harbors (48.3), Cleanaway (47.5). Veolia leads by SOV (21.3% vs Clean Harbors 9.9%, SUEZ 7.6%) but trails ENGIE by 0.6 BIS points — a narrow but real gap driven by ENGIE's stronger energy narrative.

2. Gemini gives more volume, ChatGPT gives better positioning

Engine	SOV	Avg Position	BIS
ChatGPT	21.3%	2.26	48.6
Gemini	29.0%	2.70	47.3

Gemini mentions Veolia in nearly 1 in 3 responses, but places it 0.44 positions later on average than ChatGPT does. ChatGPT mentions Veolia less frequently, but when it does, it positions it earlier in the list — where cognitive impact is highest. The BIS difference (48.6 vs 47.3) is partly explained by this positional gap. Both engines are strategically important; they require different content optimisation approaches.

3. Water Technologies and Bioenergy drive quality; Hazardous Waste drives volume

Product Line	ChatGPT BIS	ChatGPT Mentions	Gemini BIS	Gemini Mentions
Water Technologies	50.1	986	46.9	1,408
Bioenergy & Energy Efficiency	50.1	707	48.4	1,307
Hazardous Waste	47.2	1,597	46.9	1,840

Hazardous Waste generates the most raw mentions in both engines (1,597 CG / 1,840 GM), accounting for ~49% of all Veolia mentions in ChatGPT and ~40% in Gemini. But Water Technologies and Bioenergy produce higher BIS scores — the AI talks about Veolia more positively and earlier when the topic is water infrastructure or energy efficiency. The hazardous waste content base is broad but produces lower-quality positioning. This gap signals a GEO content opportunity in the waste treatment space.

4. Consideration funnel outperforms — Decision stage is the weakest link

Funnel Stage	ChatGPT BIS	ChatGPT Mentions	Gemini BIS	Gemini Mentions
Awareness	47.5	950	46.2	1,400
Consideration	49.8	1,247	48.5	1,693
Decision	48.7	1,093	47.4	1,462

Veolia performs best at Consideration stage (BIS 49.8 CG / 48.5 GM) — when buyers are comparing options and researching providers. This suggests strong brand recognition and competitive differentiation content that the AI has indexed. However, the drop at Decision stage (where buyers are ready to act and need specific validation) is a gap: Veolia loses 1.1 BIS points from Consideration to Decision in ChatGPT. Decision-stage content — case studies, certifications, contract references, outcome data — is underrepresented in the AI's knowledge base.

5. Competitive Score is high in Gemini — but Sentiment drives the BIS ceiling

The Competitive Score measures how often Veolia outranks other entities mentioned in the same response. In Gemini it reaches 0.843 — Veolia beats 84% of all co-occurring entities per response. Yet the BIS is lower in Gemini (47.3 vs 48.6). The constraint is not competitive position but mention depth: the Mention Score in Gemini (0.154) is substantially lower than in ChatGPT (0.226). Veolia appears in more Gemini responses (SOV 29% vs 21%) but is mentioned fewer times per response and with thinner contextual detail.

6. Sentiment is positive but has headroom — especially in Hazardous Waste

Both engines return positive sentiment for Veolia (CG: +0.381, GM: +0.417). This is a solid baseline — no negative narrative risk detected. But sentiment in the Hazardous Waste product line drags the average: that segment shows the lowest sentiment scores of the three lines, likely due to regulatory, compliance, and liability associations in AI-indexed content. Water Technologies and Bioenergy drive the sentiment ceiling. Improving the narrative frame around hazardous waste — focusing on safety outcomes, innovation, and circular economy angles — would lift the overall BIS by an estimated 0.5–1.0 points across both engines.

---

Summary in One Sentence

Veolia is the most visible environmental services brand in generative AI across its defined competitive set — but with a BIS of ~48, it sits in the "present but not dominant" zone, 32 points below where the AI would recommend it first, automatically and confidently. The path to dominance runs through three concrete gaps: Decision-stage content depth, Hazardous Waste sentiment elevation, and Gemini mention density.

Page 5 — Context and Methodology

---

5.1 Study Objective

Measure Veolia's visibility in ChatGPT and Gemini responses when a user asks about topics related to its three core business areas: Water Technologies & New Solutions, Hazardous Waste Treatment, and Bioenergy & Energy Efficiency.

The study covers four markets — United States, Australia, Spain, and Middle East — representing Veolia's key geographies. It benchmarks Veolia against 15 defined competitors across 36 persona profiles, producing a full-coverage map of Veolia's AI presence: where it appears, in what position, with what tone, and how it compares to the competitive field.

---

5.2 Prompt Atlas — Study Design

The Prompt Atlas generates prompts that simulate real user queries. Prompts do not ask about Veolia — they ask about the industries and challenges where Veolia operates. If the AI mentions Veolia, it does so on its own initiative.

Study Figures

Concept	Value
Total unique prompts	9,720
Neutral prompts (no "Veolia" mention)	~9,330 (~96%)
Prompts with explicit Veolia mention	~390 (~4%)
Persona × Funnel × Product Line combinations	324
Prompts per combination (avg)	30

Axis Distribution

Product Lines — 3 business areas, approximately equal coverage (~33% each):

Product Line	Prompt Atlas Instruction	Prompts
Water Technologies & New Solutions	"Water infrastructure, industrial water treatment, wastewater reuse, smart water systems, drinking water management, desalination, digital water solutions."	~3,240 (33.3%)
Hazardous Waste Treatment	"Industrial hazardous waste management, contaminated soil remediation, chemical waste treatment, secure landfill, waste incineration, compliance and traceability."	~3,240 (33.3%)
Bioenergy, Flexibility & Energy Efficiency	"Biogas and biomethane production, waste-to-energy, district heating, industrial energy efficiency, flexibility services, carbon reduction for industrial and municipal clients."	~3,240 (33.3%)

Funnels — three stages of the buyer journey:

Funnel	Instruction	Prompts
Awareness	"General industry queries. The user does NOT know any specific provider. They seek orientation: who are the main players, what solutions exist, what should they consider."	~3,240 (33.3%)
Consideration	"The user knows solutions exist and is comparing providers. Questions are evaluative: who does it better, what are the differences, which are the reference companies."	~3,240 (33.3%)
Decision	"The user is close to choosing a provider. Questions are specific and high-intent: references, certifications, case studies, contracts, outcome validation."	~3,240 (33.3%)

---

5.3 Personas — 36 Profiles Across 4 Markets

8 Archetypal Personas (×4 markets = 32 profiles)

Archetype	Brief Description
Municipal Water Infrastructure Manager	Public sector buyer responsible for water supply and wastewater systems at municipal level. Focus on compliance, long-term contracts, public accountability.
Municipal Circular Economy Manager	City-level sustainability director driving circular economy initiatives. Interested in waste reduction, resource recovery, and cross-sector partnerships.
Port Energy Infrastructure Manager	Port authority or industrial estate energy manager. Evaluating energy efficiency, on-site generation, and utility services for large logistics hubs.
Urban Energy Planning Advisor	Urban planning consultant or public agency advisor. Focus on energy transition, district-level solutions, and long-term infrastructure planning.
Industrial Sustainability Director	C-level or VP sustainability at a large industrial company. Drives ESG commitments, Scope 1/2/3 targets, and supplier decarbonisation.
Industrial Water Efficiency Manager	Operations or engineering manager at a water-intensive industrial facility (food & beverage, pharma, mining). Focused on cost reduction and regulatory compliance.
Regional Waste & Resource Recovery Director	Waste management authority or regional government official. Responsible for industrial and municipal waste streams, treatment capacity, and recovery rates.
Water & Sanitation Infrastructure Engineer	Technical project manager or procurement engineer in water utilities or engineering firms. Evaluates technical solutions, specifications, and contractor capabilities.

Markets per archetype: United States · Australia · Spain · Middle East

4 Neutral Personas (no sector profile)

Name	Market
Neutral	United States
Neutral	Australia
Neutral	Spain
Neutral	Middle East

Neutral personas provide a baseline: what does the AI say about the environmental services industry when no professional context is applied?

---

5.4 AI Engines

Engine	Responses	Brands Detected	Veolia Responses	Veolia Mentions
ChatGPT	9,374	10,633	2,066 (22.0%)	3,290
Gemini	8,826	15,739	2,736 (31.0%)	4,555
Total	18,200	—	4,802	7,845

---

5.5 Defined Competitors (15)

Competitor	Sector	Mentions ChatGPT	Mentions Gemini
SUEZ	Water & Waste	1,180	1,585
Clean Harbors	Hazardous Waste	1,160	782
Xylem	Water Technology	1,017	1,400
Cleanaway	Waste Management	437	632
Waste Management Inc.	Waste Management	198	235
Republic Services	Waste Management	323	578
REMONDIS	Waste Management	183	322
ENGIE	Energy & Utilities	347	458
Ameresco	Energy Efficiency	148	273
Aqualia	Water Services	155	197
Ecolab	Water Treatment	52	143
SAUR	Water Services	46	105
Dalkia	Energy Services	50	24
American Water	Water Utilities	23	10
Veralto	Water Analytics	6	3

Why Veolia leads

Veolia is the only defined competitor that spans all three product lines simultaneously. SUEZ focuses on water and waste; Clean Harbors on hazardous waste; ENGIE on energy. Veolia's cross-sector presence means it gets mentioned across all 9 funnel–product line combinations — while most competitors concentrate in 1–2. Its 21.3% SOV in ChatGPT and 29.0% in Gemini substantially outpace the nearest rival (Clean Harbors 9.9% CG / SUEZ 10.3% GM).

---

5.6 Sample Prompts

Water Technologies — Awareness

"Which companies operate in Spain that can manage municipal hazardous waste streams from clean points, including paints, solvents, aerosols, and batteries, with full traceability?"

Water Technologies — Consideration

"Who are the main European players with Spanish presence for hazardous waste incineration, stabilization, and secure landfill, and how do they differ in capacity?"

Hazardous Waste — Decision

"Which companies in the US can take a multi-state hazardous waste treatment contract covering solvent, paint, and lab-pack streams with strong compliance support?"

Hazardous Waste — Decision

"Which vendors offer end-to-end industrial effluent treatment and hazardous sludge disposal for food and beverage plants, including emergency response coverage?"

Hazardous Waste — Awareness

"Who are the main providers in Spain for contaminated soil remediation on former industrial land we want to redevelop into a circular economy hub?"

All examples are neutral by default — they do not mention Veolia. The AI's decision to include or exclude Veolia is entirely organic, reflecting how well Veolia's content presence aligns with each query context.

Page 6 — Competitive Benchmark: Share of Voice

---

6.1 SOV Ranking — Defined Competitors + Veolia

SOV (Share of Voice) is the percentage of total responses in which a brand appears. It measures how present a brand is across all queries, irrespective of position or tone. A brand with high SOV is one the AI mentions frequently and broadly.

Total responses: 9,374 (ChatGPT) and 8,826 (Gemini).

ChatGPT — SOV Ranking (16 entities)

#	Brand	Mentions	Responses	SOV	SBOV	BIS
1	Veolia	3,290	2,066	21.26%	22.04%	48.63
2	Clean Harbors	1,160	964	9.92%	10.28%	48.28
3	SUEZ	1,180	737	7.58%	7.86%	44.29
4	Xylem	1,017	732	7.53%	7.81%	46.92
5	Cleanaway	437	304	3.13%	3.24%	47.46
6	Republic Services	323	252	2.59%	2.69%	39.13
7	ENGIE	347	233	2.40%	2.49%	49.18
8	REMONDIS	183	161	1.66%	1.72%	40.62
9	Waste Management, Inc.	198	142	1.46%	1.51%	40.99
10	Ameresco Inc.	148	120	1.23%	1.28%	48.40
11	Aqualia	155	110	1.13%	1.17%	44.91
12	Dalkia	50	47	0.48%	0.50%	42.26
13	Ecolab	52	43	0.44%	0.46%	42.42
14	Nijhuis Saur Industries	46	39	0.40%	0.42%	42.10
15	American Water	23	19	0.20%	0.20%	41.47
16	Veralto	6	6	0.06%	0.06%	35.67

Veolia vs #2 (Clean Harbors): +11.34 percentage points gap in ChatGPT.

---

Gemini — SOV Ranking (16 entities)

#	Brand	Mentions	Responses	SOV	SBOV	BIS
1	Veolia	4,555	2,736	29.04%	31.00%	47.31
2	Xylem Inc.	1,400	1,052	11.17%	11.92%	42.82
3	SUEZ	1,585	974	10.34%	11.04%	42.07
4	Clean Harbors	782	546	5.80%	6.19%	45.89
5	Republic Services	578	468	4.97%	5.30%	41.04
6	Cleanaway	632	414	4.39%	4.69%	46.40
7	ENGIE Middle East	458	309	3.28%	3.50%	44.55
8	REMONDIS Iberia	322	277	2.94%	3.14%	38.88
9	Ameresco	273	243	2.58%	2.75%	46.79
10	Waste Management	235	178	1.89%	2.02%	38.40
11	Agbar (Aqualia)	197	158	1.68%	1.79%	40.79
12	Ecolab	143	130	1.38%	1.47%	40.88
13	Saur Group	105	90	0.96%	1.02%	41.18
14	Dalkia Middle East	24	17	0.18%	0.19%	41.76
15	American Water	10	8	0.08%	0.09%	42.38
16	Veralto	3	3	0.03%	0.03%	28.67

Veolia vs #2 (Xylem): +17.87 percentage points gap in Gemini.

---

6.2 Global Top 10 by SOV — Full Landscape

This view expands beyond the 15 defined competitors to reveal which non-competitor brands compete for AI attention in the same query space.

ChatGPT — Global Top 10

#	Brand	SOV	BIS	Mentions	Type
1	Veolia	21.26%	48.63	3,290	Target
2	Wikipedia	13.93%	36.52	2,161	Source reference
3	Clean Harbors	9.92%	48.28	1,160	Defined competitor
4	SUEZ	7.58%	44.29	1,180	Defined competitor
5	Xylem	7.53%	46.92	1,017	Defined competitor
6	Schneider Electric	5.20%	46.46	769	Adjacent (energy/water tech)
7	Siemens AG	4.55%	45.98	657	Adjacent (industrial tech)
8	Tradebe	3.96%	44.09	448	Hazardous waste (Spain)
9	Honeywell International	3.38%	42.87	419	Adjacent (industrial tech)
10	Itron	3.34%	46.57	409	Adjacent (smart water/energy)

Gemini — Global Top 10

#	Brand	SOV	BIS	Mentions	Type
1	Veolia	29.04%	47.31	4,555	Target
2	Xylem Inc.	11.17%	42.82	1,400	Defined competitor
3	SUEZ	10.34%	42.07	1,585	Defined competitor
4	Siemens AG	8.14%	42.45	997	Adjacent (industrial tech)
5	Schneider Electric	7.06%	42.22	908	Adjacent (energy/water tech)
6	ABB	6.00%	42.71	670	Adjacent (industrial automation)
7	Clean Harbors	5.80%	45.89	782	Defined competitor
8	Fluence Corporation	5.52%	45.21	628	Adjacent (water tech)
9	Republic Services	4.97%	41.04	578	Defined competitor
10	Cleanaway	4.39%	46.40	632	Defined competitor

---

6.3 Analysis

Veolia dominates by a structural margin

The SOV gap between Veolia and its nearest competitor is not marginal — it is structural. In ChatGPT, Veolia holds 21.26% against Clean Harbors' 9.92%: a 2.1× advantage over the second brand. In Gemini the lead is even more pronounced: 29.04% versus Xylem's 11.17%, a 2.6× advantage. No other brand in this study comes close to replicating Veolia's cross-sector coverage.

This lead reflects Veolia's unique position as the only defined entity that operates simultaneously across water, waste, and energy. Most competitors concentrate in one or two areas. SUEZ covers water and waste but lacks energy depth. Clean Harbors is almost entirely hazardous waste. ENGIE is primarily energy. Veolia gets mentioned across all 9 funnel × product-line combinations; competitors get mentioned in 2–4.

The #2 changes by engine — a strategic signal

In ChatGPT, the second brand is Clean Harbors (9.92%). In Gemini, it is Xylem (11.17%). This shift has a concrete explanation: ChatGPT weights hazardous waste queries heavily toward US-market specialists, and Clean Harbors is the dominant US hazardous waste brand. Gemini distributes its attention more globally and across water technology, where Xylem leads. SUEZ is close behind in both engines (7.58% CG / 10.34% GM), maintaining consistent positioning across the two.

Engine asymmetry: Gemini gives more, ChatGPT positions better

Veolia's SOV is 29.04% in Gemini vs 21.26% in ChatGPT — a 7.78 percentage point premium in Gemini. This means Gemini mentions Veolia in nearly 1 in 3 responses, while ChatGPT does so in roughly 1 in 5. However, as the BIS data shows, Gemini's higher mention rate does not translate into equivalent quality of positioning: Veolia's BIS is 48.63 in ChatGPT vs 47.31 in Gemini. The volume is higher in Gemini; the mention depth and positional impact is higher in ChatGPT.

Non-competitor intrusion: tech giants and territorial specialists

The global top 10 reveals an important dynamic: Schneider Electric, Siemens, and ABB (and Honeywell, Itron in ChatGPT) regularly appear in the same responses as Veolia. These are not Veolia's competitors in the traditional sense, but they compete for AI attention in water digitisation, industrial energy efficiency, and smart infrastructure — all areas where Veolia's content presence overlaps with broader industrial technology narratives. Their presence is a reminder that Veolia's true AI competitive landscape extends beyond the 15 defined entities.

Tradebe (3.96% SOV in ChatGPT) is a particularly notable entry: a hazardous waste specialist based in Spain that captures almost 4% of ChatGPT responses — more than ENGIE, REMONDIS, or Ameresco. This reflects the Spain market's weight in the prompt mix and signals a local competitive gap Veolia should monitor.

---

Verbatim

"For industrial hazardous waste management in the Middle East, several companies offer comprehensive solutions: 1. Veolia — operates multiple treatment facilities in the UAE, Saudi Arabia, and Qatar, offering incineration, physico-chemical treatment, and secure landfill. 2. Clean Harbors — expanding Middle East presence..."

Source: ChatGPT — Prompt: "Which companies offer hazardous waste treatment in the Gulf region?" (Industrial Sustainability Director, Decision, Hazardous Waste, Middle East) Position: 0 — Sentiment: 4/5

Veolia mentioned first with specific country coverage. The response shows how SOV leadership translates into first-mention advantage at the Decision stage — precisely where buyer intent is highest.

Page 7 — Competitive Benchmark: Brand Impact Score

---

7.1 BIS Comparative Overview

The Brand Impact Score (BIS) integrates four components into a single 0–100 indicator: Sentiment Score (positive/negative tone of mentions), Position Score (how early the brand appears in responses), Mention Score (frequency and depth of mention per response), and Competitive Score (how often the brand outranks other entities in the same response).

Veolia's BIS: 48.63 (ChatGPT) and 47.31 (Gemini).

---

ChatGPT — BIS Ranking (16 entities)

#	Brand	BIS	Position Score	Sentiment Score	Mention Score	Competitive Score
1	ENGIE	49.18	0.2803	0.4211	0.2395	0.7414
2	Veolia	48.63	0.2689	0.3806	0.2256	0.7768
3	Ameresco Inc.	48.40	0.2545	0.4090	0.2023	0.7904
4	Clean Harbors	48.28	0.2804	0.2887	0.1875	0.8630
5	Cleanaway	47.46	0.2790	0.2857	0.2100	0.7981
6	Xylem	46.92	0.2525	0.4001	0.2137	0.7114
7	Aqualia	44.91	0.2354	0.3845	0.2220	0.6350
8	SUEZ	44.29	0.2143	0.3661	0.2159	0.6494
9	Ecolab	42.42	0.1721	0.4279	0.1646	0.6268
10	Dalkia	42.26	0.1921	0.3511	0.1663	0.6521
11	Nijhuis Saur Industries	42.10	0.1937	0.3558	0.1917	0.6054
12	American Water	41.47	0.1865	0.3421	0.1791	0.6131
13	Waste Management, Inc.	40.99	0.1957	0.2386	0.1812	0.6498
14	REMONDIS	40.62	0.1973	0.2516	0.1783	0.6219
15	Republic Services, Inc.	39.13	0.1700	0.2637	0.1763	0.5767
16	Veralto	35.67	0.1861	0.0000	0.1209	0.6462

---

Gemini — BIS Ranking (16 entities)

#	Brand	BIS	Position Score	Sentiment Score	Mention Score	Competitive Score
1	Veolia	47.31	0.2136	0.4166	0.1542	0.8426
2	Ameresco	46.79	0.2059	0.4201	0.1390	0.8418
3	Cleanaway	46.40	0.1983	0.4129	0.1429	0.8334
4	Clean Harbors	45.89	0.1953	0.3704	0.1379	0.8500
5	ENGIE Middle East	44.55	0.1851	0.4042	0.1446	0.7617
6	Xylem Inc.	42.82	0.1581	0.4011	0.1236	0.7370
7	American Water	42.38	0.1899	0.3125	0.1361	0.7235
8	SUEZ	42.07	0.1645	0.3483	0.1424	0.7084
9	Dalkia Middle East	41.76	0.1853	0.3529	0.1386	0.6746
10	Saur Group	41.18	0.1509	0.3667	0.1251	0.6928
11	Republic Services	41.04	0.1473	0.3258	0.1282	0.7128
12	Ecolab	40.88	0.1231	0.4000	0.1009	0.7109
13	Agbar (Aqualia)	40.79	0.1642	0.3070	0.1263	0.6958
14	REMONDIS Iberia	38.88	0.1595	0.2589	0.1262	0.6432
15	Waste Management	38.40	0.1436	0.2669	0.1254	0.6355
16	Veralto	28.67	0.0947	0.3333	0.0711	0.2207

---

7.2 Veolia vs Competitor Pool — Competitive Analysis

The competitive analysis compares Veolia against the average of all 15 defined competitors across key metrics.

ChatGPT

Metric	Veolia	Competitors Pool Avg	Difference
BIS	48.63	41.63	+6.99
Position Score	0.2689	0.1912	+0.0777
Sentiment Score	0.3806	0.3417	+0.0389
Visibility Score	0.4134	0.3715	+0.0419

Gemini

Metric	Veolia	Competitors Pool Avg	Difference
BIS	47.31	37.02	+10.29
Position Score	0.2136	0.1074	+0.1063
Sentiment Score	0.4166	0.3611	+0.0555
Visibility Score	0.3805	0.3230	+0.0575

Veolia's BIS advantage over the pool is larger in Gemini (+10.29) than in ChatGPT (+6.99). This reflects the wider spread of the Gemini competitive field: several competitors score significantly lower in Gemini (SUEZ at 42.07, REMONDIS at 38.88, Waste Management at 38.40), while Veolia maintains a strong 47.31. In ChatGPT, the competitive field is more compressed at the top.

---

7.3 BIS Components — Veolia Strengths and Weaknesses

Strengths

Position Score — strongest relative advantage. Veolia's Position Score (0.2689 CG / 0.2136 GM) is the component where it beats the pool most decisively (+0.0777 in CG, +0.1063 in GM). This means Veolia appears earlier in responses than competitors — a critical quality signal. The AI not only mentions Veolia more often (SOV), but places it nearer the top of lists and narratives.

Competitive Score — especially strong in Gemini. At 0.8426 in Gemini, Veolia beats 84% of all co-occurring brands per response. This is the highest Competitive Score among all 16 entities in Gemini. In ChatGPT (0.7768) it ranks second, behind Clean Harbors (0.8630). A high Competitive Score means when Veolia appears alongside other brands, it tends to be positioned more favorably.

Weaknesses

Mention Score — the ceiling constraint. Veolia's Mention Score (0.2256 CG / 0.1542 GM) reflects how often and deeply the AI mentions the brand per response. It is the highest among all 16 entities in ChatGPT (0.2256), but in Gemini (0.1542) it leads by a smaller margin. The Mention Score gap between Gemini and ChatGPT (−0.0714) is the single largest component explaining the BIS difference between engines. Veolia appears in more Gemini responses (SOV 29% vs 21%), but each individual appearance contains less depth — fewer mentions per response, thinner contextual elaboration.

Sentiment Score — moderate, not leading. At 0.3806 in ChatGPT, Veolia ranks 6th out of 16 on sentiment — positive, but not the highest tone. ENGIE (0.4211), Ameresco (0.4090), Xylem (0.4001), Ecolab (0.4279), and Aqualia (0.3845) all score higher in ChatGPT sentiment. In Gemini (0.4166), Veolia ranks 2nd, behind Ameresco (0.4201). The Hazardous Waste product line is the known drag on sentiment: regulatory and compliance associations in that space produce more neutral framing than Water Technologies or Bioenergy.

---

7.4 ⚠️ ENGIE Alert — ChatGPT BIS

In ChatGPT, ENGIE (BIS 49.18) surpasses Veolia (BIS 48.63) by 0.55 points. This is the only instance across both engines where a competitor leads Veolia in BIS.

Metric	ENGIE (ChatGPT)	Veolia (ChatGPT)	ENGIE advantage
BIS	49.18	48.63	+0.55
Sentiment Score	0.4211	0.3806	+0.0405
Position Score	0.2803	0.2689	+0.0114
Mention Score	0.2395	0.2256	+0.0139
Competitive Score	0.7414	0.7768	−0.0354

ENGIE's BIS edge comes from three components simultaneously: higher sentiment (+0.04), better positioning (+0.01), and higher mention depth (+0.01). Veolia beats ENGIE only on Competitive Score. The explanation is likely topic-specific: in energy efficiency and bioenergy prompts, ENGIE's concentrated energy identity generates more pointed, enthusiastic AI responses than Veolia's broader multi-sector positioning.

In Gemini, this dynamic reverses: Veolia (47.31) leads ENGIE Middle East (44.55) by 2.76 BIS points. The ENGIE ChatGPT alert is real but narrow; it does not alter the overall picture of Veolia's leadership across both engines combined. However, it signals that in purely energy-focused query contexts, ENGIE's narrative is sharper.

---

7.5 BIS in Context

A BIS of ~48 represents solid, multi-market visibility — but not dominance. Dominant brands (BIS 70+) would be those the AI recommends first, confidently, and repeatedly in response to high-intent queries. At 48, Veolia is in the "credible reference" zone: the AI includes it consistently, often early, but rarely frames it as the singular answer.

The path from ~48 to ~70 runs through three mechanisms: increasing Mention Score (more detailed content indexed per response context), improving Sentiment Score in Hazardous Waste (more outcome-focused, less regulatory-framed content), and deepening Decision-stage materials so the AI produces specific, confident recommendations rather than general lists.

---

Verbatim

"For biogas and biomethane upgrading projects, Veolia is one of the most established operators globally. Their Bioenergy solutions cover anaerobic digestion, gas upgrading and injection, and organic waste treatment at industrial and municipal scale. ENGIE is another significant player, particularly in biomethane purchasing agreements and large-scale urban energy networks."

Source: ChatGPT — Prompt: "Which companies have the most experience operating biogas-to-biomethane projects for municipalities?" (Municipal Circular Economy Manager, Consideration, Bioenergy, Middle East) Position: 0 for Veolia — Sentiment: 4/5

In Bioenergy contexts, Veolia and ENGIE are frequently mentioned together — explaining the competitive proximity in BIS. The AI treats them as co-leaders in this domain, which is precisely why ENGIE's BIS nearly matches Veolia's in ChatGPT.

Page 8 — Competitive Benchmark: Sentiment

---

8.1 Sentiment Ranking by Brand

The Sentiment Score (−1 to +1 scale) measures the tone of AI mentions: how positively or negatively the LLM frames a brand across all responses where it appears. Scores near +1 indicate consistently enthusiastic or solution-oriented language; scores near 0 indicate neutral, informational framing; negative scores indicate critical or cautionary language.

No defined competitor in this study records a negative sentiment score. The field is uniformly positive, which reflects the nature of the query space: buyers asking about environmental services providers receive constructive, capability-focused responses. The differentiation is entirely within the positive range — from cautiously positive to actively enthusiastic.

ChatGPT — Sentiment Ranking (16 entities)

#	Brand	Sentiment Score	BIS	Mentions	Notes
1	Ecolab	0.4279	42.42	52	Small volume — score more volatile
2	ENGIE	0.4211	49.18	347	Energy narrative drives enthusiasm
3	Ameresco Inc.	0.4090	48.40	148	Energy efficiency focus
4	Xylem	0.4001	46.92	1,017	Water tech — solution-forward framing
5	Aqualia	0.3845	44.91	155	Water services specialist
6	Veolia	0.3806	48.63	3,290	Best sentiment among high-volume brands
7	SUEZ	0.3661	44.29	1,180	Water/waste — broadly positive
8	Nijhuis Saur Industries	0.3558	42.10	46	Low volume
9	Dalkia	0.3511	42.26	50	Energy services
10	American Water	0.3421	41.47	23	Low volume
11	Clean Harbors	0.2887	48.28	1,160	Hazardous waste — compliance framing
12	Cleanaway	0.2857	47.46	437	Waste management
13	Republic Services, Inc.	0.2637	39.13	323	Waste management
14	REMONDIS	0.2516	40.62	183	Waste management
15	Waste Management, Inc.	0.2386	40.99	198	Waste management
16	Veralto	0.0000	35.67	6	Negligible volume — unreliable

ChatGPT market average (15 competitors): 0.3417 Veolia vs market average (ChatGPT): +0.0389 above the pool

---

Gemini — Sentiment Ranking (16 entities)

#	Brand	Sentiment Score	BIS	Mentions	Notes
1	Ameresco	0.4201	46.79	273	Energy efficiency — strong positive framing
2	Veolia	0.4166	47.31	4,555	#1 by volume and #2 by sentiment
3	Cleanaway	0.4129	46.40	632	Australia-market prominence
4	ENGIE Middle East	0.4042	44.55	458	Energy transition narrative
5	Xylem Inc.	0.4011	42.82	1,400	Water technology focus
6	Ecolab	0.4000	40.88	143	Water treatment specialist
7	Clean Harbors	0.3704	45.89	782	Higher than ChatGPT — Gemini frames hazardous waste differently
8	Saur Group	0.3667	41.18	105	Water services
9	Dalkia Middle East	0.3529	41.76	24	Low volume
10	SUEZ	0.3483	42.07	1,585	Water/waste specialist
11	Veralto	0.3333	28.67	3	Negligible volume
12	Republic Services	0.3258	41.04	578	Waste management
13	American Water	0.3125	42.38	10	Low volume
14	Agbar (Aqualia)	0.3070	40.79	197	Water services Spain
15	Waste Management	0.2669	38.40	235	Waste management
16	REMONDIS Iberia	0.2589	38.88	322	Waste management

Gemini market average (15 competitors): 0.3611 Veolia vs market average (Gemini): +0.0555 above the pool

---

8.2 Veolia vs Market Average

Metric	ChatGPT	Gemini
Veolia Sentiment Score	0.3806	0.4166
Competitor Pool Average	0.3417	0.3611
Veolia Premium	+0.0389	+0.0555
Veolia Sentiment Rank	6th of 16	2nd of 16

Veolia scores above the competitor pool average in both engines. The premium is larger in Gemini (+0.0555) than in ChatGPT (+0.0389). In Gemini, Veolia ranks second overall — behind Ameresco by only 0.0035 points, a margin that falls within normal measurement variance.

A meaningful pattern emerges from the two tables: brands with hazardous waste as their core identity consistently score lower sentiment than water technology or energy-focused brands. Clean Harbors (0.2887 CG), Waste Management (0.2386 CG), REMONDIS (0.2516 CG), and Republic Services (0.2637 CG) all cluster at the bottom. Their sentiment scores are still positive, but materially lower. Compliance language, regulatory references, and liability framing in hazardous waste contexts produce more cautious AI narration.

Veolia spans all three domains. Its overall sentiment (0.3806 CG / 0.4166 GM) is pulled upward by its Water Technologies and Bioenergy performance, and downward by its Hazardous Waste segment. This is structurally unavoidable given Veolia's multi-sector model — but it also means targeted content investment in Hazardous Waste narrative could lift the overall score further.

---

8.3 Sentiment vs Volume — The Quality-Volume Trade-off

A key dynamic in this dataset: the brands with the highest sentiment scores tend to have low mention volumes.

Brand	ChatGPT Sentiment	ChatGPT Mentions
Ecolab	0.4279	52
ENGIE	0.4211	347
Ameresco	0.4090	148
Veolia	0.3806	3,290
SUEZ	0.3661	1,180
Clean Harbors	0.2887	1,160

Ecolab has the highest sentiment (0.4279) but only 52 mentions — the AI talks about it warmly but rarely. ENGIE scores 0.4211 with 347 mentions. Veolia, with 3,290 mentions, achieves 0.3806 — the best sentiment-at-scale combination in the competitive set. Maintaining a sentiment score above 0.38 while generating 9× more mentions than ENGIE and 63× more than Ecolab is the real achievement.

In Gemini, this pattern amplifies: Veolia has 4,555 mentions (16× more than Ameresco at 273) while scoring nearly identically on sentiment (0.4166 vs 0.4201). High volume at high sentiment is structurally harder to achieve — each additional mention creates more opportunity for the AI to introduce neutral or cautious framing. Veolia avoids this degradation.

---

8.4 Sentiment Tone Inference

Based on the Sentiment Score scale and the distribution patterns observed in comparable studies, Veolia's sentiment profile can be inferred as follows:

ChatGPT (Sentiment: 0.3806) The score of +0.381 places Veolia firmly in positive territory. Approximately 70–75% of mentions are likely positively framed (capability assertions, solution recommendations, favorable comparisons), with 25–30% neutral (informational, list-based), and negligible negative framing. This is consistent with a brand mentioned in discovery and comparison contexts where AI defaults to constructive language.

Gemini (Sentiment: 0.4166) At +0.417, Gemini's framing of Veolia is slightly warmer. The higher score reflects Gemini's tendency to include more elaborative context in its responses — when it mentions Veolia in Water Technologies or Bioenergy queries, it often adds operational detail that reads as authoritative endorsement rather than neutral listing.

---

8.5 What the AI Says About Veolia — and What It Does Not

The tone analysis reveals what the AI does well and where it is thinner:

The AI describes Veolia positively in the context of its scale ("one of the world's largest environmental services companies"), its geographic reach ("operates in 220+ countries"), and its water expertise ("over 100 years of experience in water treatment"). These are high-sentiment frames because they position Veolia as a proven, authoritative actor.

The lower sentiment in Hazardous Waste prompts reflects a different narrative pattern: the AI tends to describe Veolia in hazardous waste contexts with more procedural language ("compliant disposal", "treatment and storage", "regulatory requirements") rather than outcome-oriented language ("carbon diverted", "soil restored", "emissions reduced"). Regulatory framing is informative but tonally neutral. Outcome-focused framing generates higher sentiment scores.

This gap is the primary lever available for sentiment improvement: reframing Hazardous Waste content around environmental outcomes, circular economy recovery rates, and innovation in remediation technology — shifting the AI's indexed vocabulary from compliance language to impact language.

---

Verbatim

"Veolia's water treatment solutions for mining operations are particularly well-regarded for their closed-loop water recycling systems, which can recover over 95% of process water — significantly reducing freshwater consumption and compliance risk for mining operators in water-stressed regions."

Source: Gemini — Prompt: "Which companies offer industrial water treatment for mining operations in Australia?" (Industrial Water Efficiency Manager, Consideration, Water Technologies, Australia) Position: 0 — Sentiment: 5/5

Outcome-specific language ("recover over 95%", "water-stressed regions"), quantified impact, and dual benefit framing (cost + compliance) are the sentiment drivers here. This is the narrative profile that generates top-quartile scores.

Page 9 — Brand Ecosystem: Co-occurrence and Confusion Analysis

---

9.1 Context

Unlike a bank or foundation with a common-name prefix that creates structural confusion (e.g., "Caixa"), Veolia has a distinctive, unambiguous brand name. There is no inherent linguistic confusion risk. However, the co-occurrence data reveals a different type of signal: unexpected entities that appear alongside Veolia and raise questions about AI positioning, territorial contamination from specific markets, and competitors that the AI treats as closer alternatives than the competitive brief might suggest.

Total co-occurrence data: top 15 co-occurring entities per engine, from responses where Veolia was mentioned.

---

9.2 Co-occurrence Data

ChatGPT — Top 15 Co-occurring Entities with Veolia

#	Entity	Co-occurrences	Type	Relevance
1	SUEZ	480	Defined competitor	Expected — water/waste peer
2	Wikipedia	442	Source reference	Anomalous — AI citing a source, not a brand
3	Clean Harbors	397	Defined competitor	Expected — hazardous waste peer
4	Xylem	207	Defined competitor	Expected — water tech peer
5	Tradebe	198	Hazardous waste specialist (Spain)	Territorial — Spain market
6	Acciona	117	Infrastructure/water (Spain)	Territorial — Spain market
7	Cleanaway	109	Defined competitor	Expected — waste management peer
8	Tadweer Group	102	Waste management (UAE)	Territorial — Middle East market
9	Stericycle	94	Medical waste (US)	Sector adjacent — mixed signals
10	Montrose Environmental Group	85	Environmental consulting (US)	Sector adjacent — non-competitor
11	Republic Services	83	Defined competitor	Expected — waste peer
12	Fluence Corporation	82	Water technology startup	Adjacent — water tech
13	IDE Technologies	80	Desalination (Israel)	Sector adjacent — water tech
14	ENGIE	78	Defined competitor	Expected — energy peer
15	Urbaser	65	Urban waste management (Spain)	Territorial — Spain market

Gemini — Top 15 Co-occurring Entities with Veolia

#	Entity	Co-occurrences	Type	Relevance
1	SUEZ	630	Defined competitor	Expected — water/waste peer
2	Xylem	438	Defined competitor	Expected — water tech peer
3	Fluence Corporation	262	Water technology startup	Adjacent — water tech
4	Cleanaway	244	Defined competitor	Expected — waste management peer
5	Clean Harbors	228	Defined competitor	Expected — hazardous waste peer
6	Acciona	186	Infrastructure/water (Spain/Australia)	Territorial
7	Hydroflux EPCO	185	Water treatment (Australia)	Territorial — Australian market
8	ENGIE	155	Defined competitor	Expected — energy peer
9	Republic Services	149	Defined competitor	Expected — waste peer
10	Siemens	147	Industrial technology	Adjacent — digital water/energy
11	Clean Earth	138	Environmental services (US)	Sector adjacent — US market
12	Aquatech	129	Water treatment (US/global)	Adjacent — water tech
13	REMONDIS	125	Defined competitor	Expected — waste peer
14	Averda	125	Waste management (Middle East/Africa)	Territorial — Middle East market
15	Austrans Group	119	Waste transport (Australia)	Territorial — Australian market

---

9.3 Analysis: Four Co-occurrence Patterns

Pattern 1 — Expected competitors (no confusion risk)

SUEZ, Clean Harbors, Xylem, Cleanaway, ENGIE, Republic Services, and REMONDIS are all defined competitors that appear frequently alongside Veolia. Their co-occurrence is structurally expected: they operate in the same domains and appear in the same responses when the AI lists providers. SUEZ is Veolia's closest shadow — 480 co-occurrences in ChatGPT, 630 in Gemini — appearing more consistently with Veolia than any other entity in both engines. This confirms the AI treats Veolia and SUEZ as the canonical water/waste peer pair, which has both defensive and competitive implications.

Pattern 2 — Wikipedia as a co-occurring entity (ChatGPT only)

Wikipedia appears 442 times alongside Veolia in ChatGPT — second highest of all co-occurring entities, behind only SUEZ. This is an anomaly worth understanding: Wikipedia is not a brand or competitor, but a source that ChatGPT references or cites in the same response as Veolia. Its presence in the co-occurrence data may reflect either explicit source citations in ChatGPT's responses, or that responses mentioning Veolia tend to be the kind of encyclopedic, company-overview type that also reference Wikipedia as a source context. This entity does not appear in the Gemini top 15 at all. It has no brand confusion implications, but it is a signal that a substantial portion of Veolia's ChatGPT mentions appear in responses with an encyclopedic, reference-document framing rather than a recommendation framing.

Pattern 3 — Territorial contamination: Spain, Australia, Middle East

Three distinct territorial patterns are visible:

Spain cluster (ChatGPT): Tradebe (198 co-occurrences), Acciona (117), Urbaser (65). These are Spanish environmental/infrastructure companies that appear specifically because the study includes the Spain market with Spanish-language prompts. Tradebe is a direct hazardous waste competitor with a strong Spain presence. Acciona and Urbaser are broader infrastructure players. The AI consistently groups Veolia with these Spain-specific entities in regional query contexts. This is a real competitive dynamic in the Spain market, not an error — but it suggests that Veolia's Spanish content footprint needs to assert differentiation from Tradebe specifically.

Australia cluster (Gemini): Hydroflux EPCO (185), Austrans Group (119), and Cleanaway (244). Gemini assigns significant weight to the Australian market, likely because it indexes more Australian-language sources than ChatGPT. Hydroflux EPCO and Austrans Group are smaller, purely domestic players. Their co-occurrence with Veolia in 185 and 119 responses respectively indicates that Gemini's Australian market responses tend to list Veolia alongside local specialists — treating them as comparable alternatives at the regional level.

Middle East cluster: Tadweer Group (102 in ChatGPT), Averda (125 in Gemini). Both are Middle East waste management operators. Their presence is consistent with the study's Middle East persona coverage. Averda in particular has grown in Gemini's co-occurrence profile, suggesting Gemini's Middle East sources increasingly include this emerging regional operator alongside global players like Veolia.

Pattern 4 — Adjacent sector entities (soft confusion risk)

Stericycle (94 co-occurrences, ChatGPT): Stericycle is a US-based medical and pharmaceutical waste company — not directly in Veolia's defined competitive space. Its co-occurrence with Veolia in 94 responses suggests that in some Hazardous Waste prompts (particularly those involving laboratory, pharmaceutical, or clinical waste), the AI defaults to including Stericycle as a standard reference alongside environmental services providers. There is a mild misalignment risk: if buyers are evaluating industrial hazardous waste providers and the AI consistently includes a medical waste specialist, it can blur the relevance framing.

Montrose Environmental Group (85 co-occurrences, ChatGPT): An environmental consulting and remediation firm, primarily US-based. Its appearance alongside Veolia in contaminated soil and remediation contexts is explainable but points to a grey zone: Montrose is a consultant and contractor, not an operator at Veolia's scale, yet the AI groups them in the same response lists.

Fluence Corporation (82 CG / 262 GM): A water technology company specializing in decentralized water and wastewater treatment, particularly in emerging markets. Fluence is significantly smaller than Veolia but appears frequently in water access and emerging-market queries. In Gemini (262 co-occurrences), it ranks third overall — a signal that Gemini's sources in Water Technologies increasingly reference Fluence as an innovation-stage player alongside established operators like Veolia.

Siemens (147 in Gemini): Siemens appears as a co-occurring entity because its digital water and smart infrastructure businesses (now partly separated as Siemens Smart Infrastructure) overlap with Veolia's digital water and energy management offerings. The AI does not confuse them, but it mentions them in the same responses — usually positioning Siemens as a technology provider and Veolia as an operator.

---

9.4 No Direct Brand Confusion — But Positional Risks Exist

Unlike brand confusion scenarios where a shared name creates structural misattribution, Veolia has no equivalent risk. The brand is unique and unambiguous. However, two softer positional risks are present in the data:

Risk 1 — Scale compression in regional markets. In Spain (ChatGPT) and Australia (Gemini), the AI groups Veolia with local specialists (Tradebe, Hydroflux EPCO, Austrans Group) that operate at a fraction of Veolia's scale. This is not confusion — it is a framing limitation. The AI does not adequately differentiate between a global €42B operator and a regional SME in the same response list. Content that asserts Veolia's scale and global footprint in regional query contexts would address this.

Risk 2 — SUEZ as a permanent co-shadow. SUEZ appears alongside Veolia more than any other entity in both engines. The AI treats them as a peer pair. While competitive proximity is expected, it means that in many responses where a buyer reads Veolia, they also immediately read SUEZ. Content that draws sharp operational distinctions — technology portfolio, geographic coverage, sector specialisation — would reduce SUEZ's "free ride" in Veolia's high-SOV responses.

---

9.5 Engine Differences in Co-occurrence Patterns

Pattern	ChatGPT	Gemini
#1 co-occurring brand	SUEZ (480)	SUEZ (630)
Wikipedia present	Yes (#2, 442)	No
Australia-specific entities in top 15	No	Yes (Hydroflux, Austrans)
Spain-specific entities in top 15	Yes (Tradebe, Urbaser)	No
Tech adjacents (Siemens, Fluence)	Fluence only	Both Siemens + Fluence
Medical waste crossover	Stericycle (#9)	No

ChatGPT's co-occurrence map is more US-centric and Spain-inclusive; Gemini's is more globally distributed with stronger Australian and water-tech representation. Both engines agree on the core competitive cluster (SUEZ, Clean Harbors, Xylem, Cleanaway, ENGIE).

Page 10 — Co-occurrence Ecosystem Map

---

10.1 What the Ecosystem Reveals

Co-occurrence patterns show the AI's associative model of Veolia: which companies, technologies, and sectors appear in the same mental space when generative AI responds to queries where Veolia is mentioned. This is not a list of competitors. It is a map of how the AI contextualises Veolia — its peer group, its adjacent ecosystem, and the geographic frames in which it is understood.

The ecosystem differs meaningfully between ChatGPT and Gemini, reflecting different training source compositions and different response generation patterns.

---

10.2 ChatGPT — Full Co-occurrence Ecosystem

Top 15 Entities Co-mentioned with Veolia (ChatGPT)

#	Entity	Co-occurrences	Category	Market Anchor
1	SUEZ	480	Water/waste competitor	Global
2	Wikipedia	442	Source reference	N/A
3	Clean Harbors	397	Hazardous waste competitor	US
4	Xylem	207	Water technology competitor	Global
5	Tradebe	198	Hazardous waste specialist	Spain
6	Acciona	117	Infrastructure/water	Spain
7	Cleanaway	109	Waste management competitor	Australia
8	Tadweer Group	102	Waste management	Middle East (UAE)
9	Stericycle	94	Medical waste	US
10	Montrose Environmental Group	85	Environmental consulting	US
11	Republic Services	83	Waste management competitor	US
12	Fluence Corporation	82	Water technology (decentralized)	Global
13	IDE Technologies	80	Desalination	Israel/Global
14	ENGIE	78	Energy/utilities competitor	Global
15	Urbaser	65	Urban waste management	Spain

---

10.3 Gemini — Full Co-occurrence Ecosystem

Top 15 Entities Co-mentioned with Veolia (Gemini)

#	Entity	Co-occurrences	Category	Market Anchor
1	SUEZ	630	Water/waste competitor	Global
2	Xylem	438	Water technology competitor	Global
3	Fluence Corporation	262	Water technology (decentralized)	Global
4	Cleanaway	244	Waste management competitor	Australia
5	Clean Harbors	228	Hazardous waste competitor	US
6	Acciona	186	Infrastructure/water	Spain/Australia
7	Hydroflux EPCO	185	Water treatment (industrial)	Australia
8	ENGIE	155	Energy/utilities competitor	Global
9	Republic Services	149	Waste management competitor	US
10	Siemens	147	Industrial technology	Global
11	Clean Earth	138	Environmental services	US
12	Aquatech	129	Water treatment	US/Global
13	REMONDIS	125	Waste management competitor	Global
14	Averda	125	Waste management	Middle East/Africa
15	Austrans Group	119	Waste transport	Australia

---

10.4 Reading the Ecosystem — Four Layers

Layer 1 — The Core Peer Cluster

SUEZ, Clean Harbors, Xylem, Cleanaway, ENGIE, Republic Services form Veolia's core AI peer group. They appear in both engines with consistently high co-occurrence counts. When the AI lists environmental services providers, these entities follow Veolia most reliably.

SUEZ is the dominant shadow: 480 co-occurrences in ChatGPT, 630 in Gemini — substantially ahead of third-place Clean Harbors in both engines. The SUEZ–Veolia pair is the AI's default frame for the water and waste sector. This is a consequence of their shared history (both originate from the French infrastructure sector, both are global operators in water and waste), and it means Veolia rarely appears in a response without SUEZ also appearing. Buyers evaluating Veolia are also reading about SUEZ in the same AI response.

Xylem rises sharply in Gemini (207 → 438 co-occurrences) compared to ChatGPT, reflecting Gemini's greater emphasis on water technology innovation content. Clean Harbors drops from #3 in ChatGPT (397) to #5 in Gemini (228), consistent with ChatGPT's heavier weighting on US hazardous waste content.

Layer 2 — The Water Technology Extension

Fluence Corporation, IDE Technologies, Hydroflux EPCO, Aquatech represent an ecosystem of specialist water technology companies that the AI associates with Veolia in water infrastructure queries.

Fluence Corporation stands out: 82 co-occurrences in ChatGPT, 262 in Gemini (3rd overall). Fluence is a decentralized water treatment innovator — smaller than Veolia by two orders of magnitude, but with a strong digital/innovation content presence that Gemini indexes heavily. Its presence in Gemini's top 3 alongside Veolia in water technology queries signals an important dynamic: Gemini's water technology responses lean toward innovation-stage companies, and Fluence has positioned itself as a reference in that narrative. This is not a threat to Veolia's market position, but it is a content positioning signal — Veolia's digital water and innovation story needs to be as visible as Fluence's.

IDE Technologies (80 co-occurrences, ChatGPT only) is a desalination specialist, most prominent in Middle East water infrastructure queries. Its presence confirms that Veolia's desalination and water infrastructure prompts for the Middle East market generate a specific peer cluster.

Aquatech (129, Gemini) and Hydroflux EPCO (185, Gemini) are both industrial water treatment specialists — the former US-based, the latter Australian. Their strong Gemini presence alongside Veolia confirms that Gemini's water technology responses for industrial customers are more localized and specialist than ChatGPT's.

Layer 3 — Industrial Technology Adjacents

Siemens (147, Gemini), Acciona (186 GM / 117 CG) represent the industrial technology and infrastructure conglomerate layer.

Siemens appears only in Gemini's top 15 (147 co-occurrences) — entirely absent from ChatGPT's. Gemini sources contain significant industrial technology content that covers Siemens' water and energy management divisions alongside Veolia's operations. This is not confusion — the AI presents them as complementary (Siemens as technology provider, Veolia as operator) — but it is a market signal: in the AI's Gemini model, Veolia's digital infrastructure narrative is contextualized alongside Siemens and Schneider Electric, not just alongside SUEZ and Clean Harbors.

Acciona appears in both engines (117 CG, 186 GM). It is an unusual presence: a Spanish infrastructure conglomerate with water and environment divisions, primarily known for construction and renewable energy. Its co-occurrence reflects Veolia's strong Spain market presence — many Spain-specific prompts generate Acciona alongside Veolia in water infrastructure and urban services contexts.

Layer 4 — Geographic Specialists (Market-Specific Entities)

Several entities appear specifically because the study covers four markets, and the AI's sources are market-weighted.

Entity	Co-occurrences	Market	Profile
Tadweer Group	102 (CG)	Middle East (UAE)	Abu Dhabi Waste Management Centre
Averda	125 (GM)	Middle East/Africa	Regional waste management operator
Tradebe	198 (CG)	Spain	Hazardous waste — Spain's leading independent
Urbaser	65 (CG)	Spain	Urban solid waste management
Hydroflux EPCO	185 (GM)	Australia	Industrial water treatment
Austrans Group	119 (GM)	Australia	Waste transport and logistics
Clean Earth	138 (GM)	US	Specialty waste and contaminated materials
Stericycle	94 (CG)	US	Medical/pharmaceutical waste

These entities confirm that the AI tailors its competitive landscape to geography. When a Middle East persona asks about waste management, Tadweer Group and Averda appear alongside Veolia — not REMONDIS or Cleanaway. When an Australian persona asks about industrial water, Hydroflux EPCO appears alongside Veolia — not IDE Technologies. The AI's associative model is market-sensitive, not just sector-sensitive.

---

10.5 Engine Differences — What Each AI "Thinks" About Veolia

Dimension	ChatGPT Ecosystem	Gemini Ecosystem
Top co-occurring brand	SUEZ (480)	SUEZ (630)
Water tech layer	Fluence, IDE	Fluence, Hydroflux, Aquatech
Energy adjacents	ENGIE only	ENGIE + Siemens
Spain market presence	Tradebe, Acciona, Urbaser	Acciona only
Australia market presence	Cleanaway only	Cleanaway, Hydroflux, Austrans
Middle East presence	Tadweer	Averda
Source references	Wikipedia (442)	Not present
US specialists	Stericycle, Montrose	Clean Earth
Innovation brands	Fluence, IDE	Fluence, Aquatech, Hydroflux

ChatGPT builds a more US-centric, Spain-weighted ecosystem. Its co-occurrence map weights hazardous waste (Clean Harbors #3, Stericycle #9, Montrose #10, Tradebe #5) and Spain-specific players heavily. The Wikipedia appearance (442 co-occurrences) suggests ChatGPT frequently produces reference-type responses about Veolia alongside encyclopedia-style citations.

Gemini builds a more globally distributed, water-technology-forward ecosystem. It assigns more weight to the Australian market (Hydroflux, Austrans), Gemini-indexed innovation companies (Fluence, Aquatech), and industrial conglomerates (Siemens). Gemini's Veolia ecosystem is richer in water technology specialists and more geographically diverse.

The consistent element across both engines is the SUEZ dominance at #1 and the presence of the five core defined competitors (Clean Harbors, Xylem, Cleanaway, ENGIE, Republic Services) in both top 15 lists. The peripheral layers diverge significantly.

---

10.6 Strategic Implications

1. SUEZ pairing is structural — differentiation content is the lever. Since SUEZ appears in 480–630 responses alongside Veolia, buyers reading AI responses about Veolia will also read about SUEZ in the same response. Content that explicitly distinguishes Veolia's capabilities (technology portfolio breadth, geographic coverage, integration of water-waste-energy) will reduce the cognitive equivalence the AI currently implies between the two brands.

2. Fluence's Gemini rise signals a content gap in digital water. With 262 Gemini co-occurrences, Fluence Corporation now appears in more Gemini responses alongside Veolia than Clean Harbors does. This is a content market signal: Gemini's AI model for water technology innovation has indexed Fluence-originated content heavily. Veolia's digital water platform and innovation pipeline needs equivalent content depth and visibility in the sources Gemini prioritizes.

3. Market-specific content unlocks geographic differentiation. The presence of Tradebe (Spain), Tadweer/Averda (Middle East), and Hydroflux/Austrans (Australia) shows the AI is sensitive to geography. Veolia's local market content — case studies, technical references, local partnerships — directly influences which peer group the AI places Veolia in for each market context. Investing in Spain-, Middle East-, and Australia-specific structured content will shift the local co-occurrence profile toward Veolia-favorable associations.

4. Siemens and Schneider Electric are the invisible competitors. They appear as co-occurrences but not in the defined competitor list. Yet in Gemini, Siemens co-occurs with Veolia more than REMONDIS or Waste Management (both defined competitors). The AI increasingly frames industrial water and energy management as a technology challenge — and includes industrial technology giants in that frame. Veolia's positioning as a technology-enabled operator, not just a service provider, needs to be explicit enough that the AI distinguishes Veolia's role from a technology vendor's.

---

Verbatim

"In Australia, the leading providers for industrial wastewater treatment and resource recovery include: 1. Veolia Water Technologies — offers full-cycle industrial water management with strong presence in mining, food and beverage, and municipal sectors. 2. Hydroflux EPCO — specialist in industrial water and wastewater for process industries. 3. Cleanaway — broad waste and resource recovery with growing water services division..."

Source: Gemini — Prompt: "Which companies offer industrial wastewater treatment for process industries in Australia?" (Industrial Water Efficiency Manager, Consideration, Water Technologies, Australia) Position: 0 for Veolia — Sentiment: 4/5

Veolia leads the response, but Hydroflux EPCO (185 co-occurrences) appears second — confirming the Australian water technology ecosystem pattern. The response demonstrates how a globally dominant brand navigates local market contexts: Veolia holds the #1 position but the AI immediately introduces a local specialist as a credible alternative.

Page 11 — Perception by Product Line

---

11.1 The Volume–Quality Inversion

The Prompt Atlas distributes prompts across three product lines that mirror Veolia's strategic portfolio: Water Technologies & New Solutions, Hazardous Waste Treatment, and Bioenergy & Energy Efficiency. Mention counts are not proportional to BIS scores — Hazardous Waste dominates volume while Water Technologies and Bioenergy lead on quality. This inversion is the central finding of this section.

ChatGPT — Perception by Product Line

Product Line	Mentions	Responses	Sentiment	Avg Position	BIS
Water Technologies	986	630	+0.432	0.281	50.11
Bioenergy & Energy Efficiency	707	404	0.428	0.293	50.08
Hazardous Waste	1,597	1,032	0.331	0.252	47.15

Gemini — Perception by Product Line

Product Line	Mentions	Responses	Sentiment	Avg Position	BIS
Bioenergy & Energy Efficiency	1,307	734	0.418	0.236	48.39
Water Technologies	1,408	952	0.436	0.202	46.90
Hazardous Waste	1,840	1,050	0.398	0.209	46.93

---

11.2 Hazardous Waste: The Volume Paradox

Hazardous Waste generates more raw mentions than the other two lines combined in ChatGPT (1,597 vs 986 + 707 = 1,693 — nearly matched) and substantially more in Gemini (1,840 vs 1,307 + 1,408 = 2,715). It accounts for roughly 49% of all Veolia mentions in ChatGPT and 40% in Gemini.

Yet it produces the lowest BIS in ChatGPT (47.15 — 2.96 points below Water Technologies) and near-bottom BIS in Gemini (46.93). The sentiment gap explains part of this: Hazardous Waste sentiment is +0.331 in ChatGPT versus +0.432 for Water Technologies — a 10-point normalized gap in how positively the AI frames Veolia in this context.

Why? Hazardous waste content inherently carries regulatory, compliance, liability, and risk associations. When the AI generates a response about industrial hazardous waste management, it frames providers through the lens of regulatory adherence and risk containment — not innovation or leadership. Veolia appears frequently, but in a more cautious, obligation-framed narrative. The volume is high because Veolia genuinely dominates this market and AI training data reflects that dominance. The quality is lower because the content ecosystem around hazardous waste is compliance-heavy rather than aspiration-heavy.

---

11.3 Water Technologies: Authority in a Specialist Domain

Water Technologies produces the highest BIS in ChatGPT (50.11) despite ranking second in mentions volume. The 0.432 sentiment score is the strongest of the three lines in ChatGPT. The 0.281 average position score indicates Veolia appears earlier in water-related responses than in waste or energy ones.

The AI attributes specific capabilities to Veolia in this domain: - Industrial water treatment and reuse (municipal and mining applications) - Membrane filtration, ultrafiltration, reverse osmosis systems - Desalination and zero-liquid-discharge solutions - Water recycling in industrial processes

This level of technical specificity generates higher BIS because the AI is citing Veolia as a named authority on concrete technologies — not just as a provider of a general service category. The mention depth is richer.

In Gemini, Water Technologies drops to second place (BIS 46.90), just 0.03 BIS points below Hazardous Waste. The sentiment advantage (+0.436) remains, but the position score is lower (0.202 vs 0.252 in ChatGPT). Gemini places Veolia less prominently in water-technology responses compared to how ChatGPT does, suggesting weaker indexing of Veolia's water-specific content in Gemini's training corpus.

---

11.4 Bioenergy & Energy Efficiency: Small Volume, Disproportionate Quality

Bioenergy & Energy Efficiency has the fewest ChatGPT mentions (707) and the fewest ChatGPT responses (404) — roughly half the volume of Water Technologies and less than half of Hazardous Waste. Yet its BIS (50.08) nearly matches Water Technologies (50.11), and its position score (0.293) is actually the highest of the three lines in ChatGPT.

This is the clearest efficiency signal in the study: Veolia gets maximum BIS impact per mention in Bioenergy. When the AI talks about Veolia in an energy efficiency context, it places it prominently and with positive framing. The energy transition narrative — district heating, waste-to-energy, industrial decarbonization — plays well in AI responses because it aligns with topics that attract high-quality, authoritative content online.

In Gemini, Bioenergy is the clear top performer (BIS 48.39 — 1.46 points above Water Technologies and 1.46 points above Hazardous Waste). This is the only product line where Gemini delivers better BIS than ChatGPT, suggesting that Veolia's energy efficiency content performs particularly well in Gemini's training sources.

---

11.5 What Attributes Does the AI Associate with Veolia by Domain?

The attribute detection data (ChatGPT) shows how the AI frames Veolia across all product lines:

Attribute	Detections	Avg Sentiment (1–5)
Services	634	3.69
Sustainability	484	4.00
Technology	430	4.00
Compliance History	427	3.95
Innovation	199	4.01
Experience	124	4.00
Quality	118	4.03
Integration	117	3.88
Transparency	77	3.97
Safety	64	4.00

Services leads in raw detections (634) but has the lowest sentiment (3.69/5) — a generic framing that doesn't differentiate. Innovation and Quality produce the strongest sentiment scores (4.01 and 4.03), but are detected far less frequently (199 and 118). Compliance History has 427 detections — concentrated in Hazardous Waste responses — and reflects the regulatory framing discussed above.

The strategic implication: the AI knows what Veolia does (Services: 634 detections) but attributes less quality to it than it could. Innovation and Quality are under-surfaced relative to their sentiment impact.

---

11.6 Strategic Implication

The three-line picture is clear:

Product Line	Position	Opportunity
Water Technologies	Strong in ChatGPT, weaker in Gemini	Replicate ChatGPT content quality in Gemini-indexed sources
Bioenergy & Energy Efficiency	Highest efficiency (BIS per mention), Gemini leader	Scale content volume — quality already high, volume is the constraint
Hazardous Waste	Highest volume, lowest quality	Shift framing from compliance to outcomes: circular economy, safety innovation, zero-waste targets

Hazardous Waste is not underperforming because Veolia is invisible — it's underperforming because the AI's narrative frame around hazardous waste inherently depresses sentiment. The content intervention needed here is not quantity but angle: case studies, innovation milestones, safety outcome data, and circular economy positioning that repositions Veolia from "compliant handler" to "circular resource manager."

Page 12 — Territorial Bias

---

12.1 Market Structure of the Study

The Prompt Atlas covers four markets with 9 persona profiles each (36 total), generating prompts in English with market-specific contextual framing. The four markets are: United States, Australia, Spain, and Middle East. Each market reflects a different competitive landscape, content availability profile, and Veolia commercial footprint.

The territorial analysis groups personas by country suffix and calculates per-market averages for BIS, sentiment, and response volume. All figures below come from the persona-level data.

---

12.2 Market Comparison Table

ChatGPT — Veolia Visibility by Market (average of 9 personas per market)

Market	Avg Responses	Avg BIS	Avg Sentiment	Total Responses
Middle East	68.0	51.24	+0.405	612
Australia	56.3	47.61	+0.371	507
Spain	57.4	47.40	+0.363	517
United States	47.8	46.59	+0.356	430

Gemini — Veolia Visibility by Market (average of 9 personas per market)

Market	Avg Responses	Avg BIS	Avg Sentiment	Total Responses
Middle East	89.8	48.47	+0.412	808
Australia	89.2	47.63	+0.437	803
Spain	63.4	46.37	+0.390	571
United States	61.6	45.53	+0.413	554

---

12.3 The Middle East Leads Both Engines

The Middle East is Veolia's strongest market in generative AI across both ChatGPT and Gemini. In ChatGPT, the average BIS for Middle East personas (51.24) is 2.65 points above Australia and 4.65 points above the United States. In Gemini, the Middle East leads with 48.47 — 0.84 points ahead of Australia and 2.94 points ahead of the US.

The Middle East advantage operates through two mechanisms:

1. Concentration of large-scale infrastructure projects. The region's desalination, water management, and industrial energy infrastructure — including NEOM, Abu Dhabi utilities, and Gulf industrial zones — generates high volumes of AI-indexed content where Veolia is named as a key provider. When a Water & Sanitation Infrastructure Engineer in the Middle East persona queries the AI about large-scale water projects, Veolia appears with specific project references.

2. Reduced competitive fragmentation. The Middle East market has fewer local competitors indexed by the AI. Where ChatGPT might list 8–10 comparable companies for a US query, Middle East queries return shorter lists — and Veolia's prominence increases when it competes with fewer alternatives in the response. The Port Energy Infrastructure Manager - Middle East (BIS 50.26 ChatGPT / 51.13 Gemini) and Industrial Water Efficiency Manager - Middle East (BIS 53.76 ChatGPT) reflect this dynamic clearly.

3. Sentiment quality. Middle East ChatGPT sentiment (+0.405) is the highest of all markets, likely because Veolia's Gulf projects are cited in positive, capacity-building narratives — energy security, water sovereignty, infrastructure modernization — rather than the compliance-heavy framing that drags down Hazardous Waste sentiment in US or Spain contexts.

---

12.4 The United States: Most Competitive, Lowest BIS

The US is Veolia's weakest market by average BIS: 46.59 in ChatGPT and 45.53 in Gemini — roughly 4.7 and 3.0 points below the Middle East respectively. This is counterintuitive given Veolia's significant US operations, but the data explains it: the US market has the most competitive AI landscape in this study.

US queries in the water, waste, and energy domains return responses populated by strong domestic brands: Clean Harbors, Waste Management Inc., Republic Services, American Water, Ameresco, and ENGIE North America. All have high English-language content volumes, strong Wikipedia and industry report presence, and established AI citation patterns. Veolia is visible — but it is one of many, not a dominant reference.

The Urban Energy Planning Advisor - United States (BIS 45.86 ChatGPT / 42.53 Gemini) is the lowest-performing single persona in both engines, reflecting that in US energy advisory contexts, domestic companies (Ameresco, ENGIE) outrank Veolia more consistently.

Response volume is also lowest in the US (430 ChatGPT / 554 Gemini total responses), indicating that US personas generate fewer Veolia mentions — not because the AI doesn't know Veolia, but because the competitive response density dilutes its share.

---

12.5 Australia: Strong Positioning, High Sentiment in Gemini

Australia ranks second in both engines (BIS 47.61 ChatGPT / 47.63 Gemini). What makes Australia interesting is its Gemini sentiment score (+0.437) — the highest of all four markets in Gemini, and substantially above the Middle East (+0.412) and US (+0.413) in the same engine.

Veolia's Australian operations — particularly through its Cleanaway-adjacent waste services and water management contracts — are well-represented in English-language Australian industry content. Gemini, which appears to weight English-speaking market sources more evenly across geographies, captures this positively. The Industrial Water Efficiency Manager - Australia (BIS 51.28 Gemini) is one of the top-performing individual personas in the entire Gemini dataset.

The Australian market has the additional characteristic of a clear competitor (Cleanaway) that appears frequently in co-occurrence data — but as a defined competitor, not a dominant one. Veolia's positioning relative to Cleanaway remains favorable.

---

12.6 Spain: Mid-Table with Territorial Fragmentation

Spain ranks third in ChatGPT (BIS 47.40) and third in Gemini (BIS 46.37). The Gemini gap is larger — Spain trails Australia by 1.26 BIS points in Gemini versus 0.21 points in ChatGPT. This divergence suggests that Gemini indexes less high-quality Veolia content from Spanish sources compared to ChatGPT.

Spain shows the clearest evidence of territorial fragmentation. Local competitors — Tradebe (hazardous waste), Acciona (water/infrastructure), Agbar/Aqualia (water utilities), and Urbaser (urban waste) — all appear in Veolia's co-occurrence data for Spanish contexts. These competitors have high Spanish-language content volumes that the AI indexes, creating a more crowded positioning landscape.

The Regional Waste & Resource Recovery Director - Spain (BIS 44.39 ChatGPT) is among the bottom performers in the study, reflecting that Spanish waste management queries return a particularly competitive response landscape where Tradebe, Urbaser, and FCC Medio Ambiente dilute Veolia's share. Port Energy Infrastructure Manager - Spain (BIS 44.92 ChatGPT) similarly underperforms, as Spanish energy infrastructure queries surface Endesa, Naturgy, and Acciona before Veolia in many responses.

Spanish-language content for Veolia may also be thinner or less structured than English-language equivalents, contributing to lower AI indexability.

---

12.7 ChatGPT vs Gemini: Engine Divergence by Market

Market	ChatGPT BIS	Gemini BIS	Gap (CG–GM)
Middle East	51.24	48.47	+2.77
Australia	47.61	47.63	−0.02
Spain	47.40	46.37	+1.03
United States	46.59	45.53	+1.06

ChatGPT outperforms Gemini in every market except Australia (where the two engines are effectively tied at 47.6). The Middle East shows the largest ChatGPT advantage (2.77 points), suggesting that Veolia's Gulf project references are particularly well-indexed in ChatGPT's training data.

Australia is the only market where Gemini sentiment (+0.437) meaningfully exceeds ChatGPT sentiment (+0.371), indicating that Gemini's Australian-sourced content carries a more positive framing of Veolia in that market.

---

12.8 Strategic Implication by Market

Market	Priority	Action
Middle East	Defend and expand	Content depth: project case studies, water security outcomes, Gulf infrastructure references
Australia	Leverage Gemini strength	Amplify English-language Australian content for Gemini indexing; reinforce water treatment and waste recovery positioning
Spain	Differentiate from locals	Spanish-language thought leadership to counter Tradebe, Acciona, Aqualia; separate waste and water narratives
United States	Improve competitive share	Energy transition and water technology content to compete with Ameresco and ENGIE; decision-stage case studies with US references

Page 13 — Persona Analysis

---

13.1 Full Persona Table

The study uses 36 persona profiles: 9 archetypes × 4 markets. Each persona has specific professional framing that shapes the type of query it generates. Tables below show all 36 personas sorted by BIS descending within each engine.

ChatGPT — All 36 Personas

Persona	Market	Responses	Sentiment	BIS
Industrial Water Efficiency Manager	Middle East	91	+0.455	53.76
Industrial Sustainability Director	Middle East	71	+0.431	52.94
Municipal Circular Economy Manager	Middle East	71	+0.371	51.92
Water & Sanitation Infrastructure Engineer	Middle East	59	+0.394	50.86
Regional Waste & Resource Recovery Director	Middle East	77	+0.400	50.65
Urban Energy Planning Advisor	Australia	53	+0.389	50.15
Neutral	Middle East	74	+0.402	50.30
Municipal Water Infrastructure Manager	Middle East	57	+0.396	50.28
Port Energy Infrastructure Manager	Middle East	46	+0.377	50.26
Urban Energy Planning Advisor	Middle East	66	+0.419	50.21
Industrial Water Efficiency Manager	United States	78	+0.440	49.95
Industrial Water Efficiency Manager	Spain	86	+0.431	49.91
Industrial Sustainability Director	United States	61	+0.370	48.85
Municipal Water Infrastructure Manager	Spain	52	+0.352	48.88
Neutral	Spain	47	+0.383	48.28
Regional Waste & Resource Recovery Director	Australia	57	+0.393	48.12
Municipal Circular Economy Manager	Spain	50	+0.369	48.06
Industrial Water Efficiency Manager	Australia	85	+0.404	49.12
Industrial Sustainability Director	Spain	59	+0.383	49.19
Industrial Sustainability Director	Australia	54	+0.367	49.07
Water & Sanitation Infrastructure Engineer	Australia	56	+0.387	47.95
Neutral	Australia	52	+0.364	46.94
Water & Sanitation Infrastructure Engineer	United States	45	+0.374	46.93
Municipal Water Infrastructure Manager	Australia	55	+0.349	46.60
Regional Waste & Resource Recovery Director	United States	57	+0.386	46.09
Municipal Water Infrastructure Manager	United States	39	+0.310	46.67
Water & Sanitation Infrastructure Engineer	Spain	57	+0.311	46.75
Urban Energy Planning Advisor	Spain	59	+0.371	46.22
Neutral	United States	36	+0.368	45.61
Urban Energy Planning Advisor	United States	29	+0.322	45.86
Municipal Circular Economy Manager	Australia	58	+0.365	45.78
Regional Waste & Resource Recovery Director	Spain	54	+0.343	44.39
Port Energy Infrastructure Manager	Spain	53	+0.321	44.92
Municipal Circular Economy Manager	United States	50	+0.310	44.68
Port Energy Infrastructure Manager	Australia	37	+0.317	44.78
Port Energy Infrastructure Manager	United States	35	+0.321	44.71

Gemini — All 36 Personas

Persona	Market	Responses	Sentiment	BIS
Industrial Water Efficiency Manager	Australia	114	+0.478	51.28
Port Energy Infrastructure Manager	Middle East	56	+0.438	51.13
Industrial Water Efficiency Manager	Middle East	149	+0.429	49.67
Water & Sanitation Infrastructure Engineer	Middle East	93	+0.422	49.70
Industrial Sustainability Director	Middle East	80	+0.426	48.85
Urban Energy Planning Advisor	Middle East	85	+0.423	48.66
Water & Sanitation Infrastructure Engineer	Australia	83	+0.448	48.17
Municipal Circular Economy Manager	Australia	106	+0.436	48.03
Municipal Circular Economy Manager	Middle East	90	+0.383	48.34
Industrial Sustainability Director	Australia	74	+0.442	47.84
Municipal Water Infrastructure Manager	Australia	88	+0.448	47.42
Port Energy Infrastructure Manager	Australia	71	+0.441	47.37
Port Energy Infrastructure Manager	Spain	54	+0.335	47.37
Industrial Water Efficiency Manager	Spain	88	+0.435	47.98
Municipal Water Infrastructure Manager	Middle East	81	+0.395	47.38
Industrial Sustainability Director	United States	68	+0.429	47.35
Urban Energy Planning Advisor	Spain	67	+0.431	47.13
Industrial Water Efficiency Manager	United States	111	+0.430	47.56
Regional Waste & Resource Recovery Director	Australia	96	+0.428	46.91
Urban Energy Planning Advisor	Australia	102	+0.405	46.86
Port Energy Infrastructure Manager	United States	39	+0.402	46.62
Industrial Sustainability Director	Spain	50	+0.398	46.98
Neutral	Spain	53	+0.434	46.38
Regional Waste & Resource Recovery Director	Middle East	102	+0.390	46.35
Water & Sanitation Infrastructure Engineer	Spain	76	+0.360	46.26
Neutral	Middle East	72	+0.403	46.14
Regional Waste & Resource Recovery Director	United States	82	+0.445	45.72
Municipal Circular Economy Manager	Spain	60	+0.329	45.57
Municipal Water Infrastructure Manager	Spain	63	+0.368	45.54
Municipal Circular Economy Manager	United States	51	+0.444	45.22
Municipal Water Infrastructure Manager	United States	53	+0.382	45.06
Neutral	Australia	69	+0.409	44.84
Neutral	United States	44	+0.412	44.86
Water & Sanitation Infrastructure Engineer	United States	57	+0.388	44.84
Regional Waste & Resource Recovery Director	Spain	60	+0.421	44.10
Urban Energy Planning Advisor	United States	49	+0.388	42.53

---

13.2 Best and Worst Performing Profiles

ChatGPT

Best persona: Industrial Water Efficiency Manager – Middle East (BIS 53.76, sentiment +0.455, 91 responses). This profile queries the AI as a water treatment procurement manager in a Gulf context — a combination that maximizes every BIS component: the AI has rich Veolia project references for Gulf water infrastructure, positions Veolia prominently (often first or second), and frames it with high positive sentiment tied to water security and technology innovation.

Worst persona: Port Energy Infrastructure Manager – United States (BIS 44.71, sentiment +0.321, 35 responses). Port and energy infrastructure in a US context generates a competitive response landscape (ENGIE, Ameresco, LM Wind Power, Wärtsilä) where Veolia is neither the natural first reference nor the dominant specialist. It appears but rarely leads.

Gemini

Best persona: Industrial Water Efficiency Manager – Australia (BIS 51.28, sentiment +0.478, 114 responses). Australian industrial water queries — particularly for mining and resource extraction contexts — appear to have strong Veolia content indexed by Gemini, likely from Australian industry publications and water authority reports.

Worst persona: Urban Energy Planning Advisor – United States (BIS 42.53, sentiment +0.388, 49 responses). US urban energy planning queries center on domestic energy companies and municipal utilities. Veolia's energy advisory positioning in the US is less established in AI-indexed content compared to its water or waste presence. This is a 6-point BIS gap below the US Industrial Water Efficiency Manager in Gemini.

---

13.3 Technical vs Managerial Profiles

Grouping the 8 archetypes by profile type reveals a consistent pattern:

Technical profiles (Water & Sanitation Infrastructure Engineer, Municipal Water Infrastructure Manager, Port Energy Infrastructure Manager) tend to generate more specific, infrastructure-focused queries. Veolia's visibility in these contexts depends heavily on whether AI training data includes detailed technical content about Veolia's infrastructure projects.

Managerial profiles (Industrial Sustainability Director, Industrial Water Efficiency Manager, Municipal Circular Economy Manager, Regional Waste & Resource Recovery Director, Urban Energy Planning Advisor) generate queries framed around vendor selection, benchmarking, and strategy — contexts where Veolia's brand recognition and competitive positioning are more directly surfaced.

Average BIS by Profile Type (ChatGPT, cross-market mean)

Profile Type	Archetype	Avg BIS (4 markets)
Managerial	Industrial Water Efficiency Manager	50.73
Managerial	Industrial Sustainability Director	50.01
Neutral	Neutral	47.66
Technical	Water & Sanitation Infrastructure Engineer	47.86
Managerial	Municipal Circular Economy Manager	47.60
Managerial	Regional Waste & Resource Recovery Director	47.46
Technical	Municipal Water Infrastructure Manager	48.11
Managerial	Urban Energy Planning Advisor	48.11
Technical	Port Energy Infrastructure Manager	46.17

Managerial profiles generate higher average BIS than technical ones in ChatGPT. The exception is Municipal Water Infrastructure Manager, which performs at managerial levels (48.11) because water infrastructure procurement queries are well-served by Veolia's documented project portfolio. Port Energy Infrastructure Manager is the weakest archetype overall, averaging 46.17 across all markets — suggesting Veolia's port and terminal energy presence is underrepresented in AI training data.

---

13.4 The Neutral Baseline

The Neutral persona profiles (one per market, no specialist framing) serve as the baseline — they query the AI without professional context or domain restriction. Their average BIS across markets:

Market	Neutral BIS (ChatGPT)	Neutral BIS (Gemini)
Middle East	50.30	46.14
Spain	48.28	46.38
Australia	46.94	44.84
United States	45.61	44.86

Key finding: Specialized managerial profiles outperform neutral baselines in ChatGPT in every market — suggesting that professional context helps the AI locate and cite Veolia more precisely. However, in Gemini, neutral personas sometimes approach or match specialized personas, indicating that Gemini responds less differentially to persona framing.

The Neutral-United States (BIS 45.61 ChatGPT) is the second-lowest neutral score, reflecting that generic environmental/sustainability queries in the US are answered with a crowded competitor mix where Veolia does not automatically lead.

---

13.5 Which Archetypes Consistently Outperform Across Markets?

Two archetypes consistently deliver above-average BIS across all four markets in ChatGPT:

Industrial Water Efficiency Manager — average BIS 50.73 (ChatGPT). This archetype outperforms in every market: ME (53.76), US (49.95), ES (49.91), AU (49.12). Water efficiency queries in industrial contexts reliably surface Veolia as a key reference, regardless of geography. It is Veolia's strongest archetype by a 0.72-point margin over the second-best.

Industrial Sustainability Director — average BIS 50.01 (ChatGPT). This archetype queries from a corporate sustainability strategy perspective — ESG reporting, decarbonization, circular economy goals. Veolia's sustainability narrative (sustainability: 484 attribute detections, sentiment 4.00/5 in ChatGPT attributes data) supports strong visibility here. Performance is consistent: ME (52.94), US (48.85), ES (49.19), AU (49.07).

Port Energy Infrastructure Manager is the consistently weakest archetype — the only one that averages below 46.0 across markets in ChatGPT (average 46.17), with all four market variants scoring below 45.0 except Middle East (50.26). Port-specific energy content for Veolia is sparse in AI training data.

Page 14 — Funnel Analysis

---

14.1 Three Stages, Three Types of Intent

The Prompt Atlas structures prompts across three buyer journey stages, each reflecting a distinct information need:

Funnel Stage	User Intent	Example Query Type
Awareness	Exploring the problem space, not yet comparing providers	"Which companies provide industrial water treatment for mining?"
Consideration	Comparing specific providers against alternatives	"Compare Veolia, SUEZ, and Xylem for water recycling in semiconductor manufacturing"
Decision	Ready to act, seeking validation and specifics	"What certifications does Veolia hold for hazardous waste treatment in the US?"

---

14.2 Funnel Results by Engine

ChatGPT — Performance by Funnel Stage

Funnel Stage	Mentions	Responses	Sentiment	Avg Position	BIS
Awareness	950	737	+0.384	0.249	47.48
Consideration	1,247	684	+0.392	0.284	49.78
Decision	1,093	645	+0.364	0.277	48.72

Gemini — Performance by Funnel Stage

Funnel Stage	Mentions	Responses	Sentiment	Avg Position	BIS
Awareness	1,400	999	+0.421	0.191	46.19
Consideration	1,693	909	+0.410	0.234	48.50
Decision	1,462	828	+0.419	0.218	47.36

---

14.3 Consideration Leads Both Engines

Consideration is the strongest funnel stage for Veolia in both ChatGPT and Gemini. It delivers the highest BIS (49.78 ChatGPT / 48.50 Gemini), the highest mention volume (1,247 ChatGPT / 1,693 Gemini), and the best position score in ChatGPT (0.284) and Gemini (0.234).

Why does Consideration outperform? The framing of Consideration prompts asks the AI to explicitly compare providers — "Which companies offer X for Y industry?" or "Compare providers of Z service." In these responses, the AI generates structured lists where Veolia's name recognition, breadth of service lines, and market leadership cause it to appear consistently and prominently. Veolia competes across three distinct domains (water, waste, energy), which means it qualifies for a wider range of comparative queries than single-domain competitors like Clean Harbors (waste only) or Xylem (water only). The breadth creates a structural Consideration advantage: Veolia appears in three times more category comparisons than mono-domain rivals.

The 2.30-point BIS advantage of Consideration over Awareness in ChatGPT (49.78 vs 47.48) is the largest inter-stage gap in the study. It confirms that Veolia's AI presence is strongest when buyers are actively evaluating vendors — not when they are discovering the problem space.

---

14.4 Decision Stage: Consistent But Not Optimal

Decision stage produces the second-highest BIS in both engines (48.72 ChatGPT / 47.36 Gemini), outperforming Awareness but trailing Consideration. This pattern is consistent across engines, suggesting a real structural dynamic rather than engine-specific noise.

The Decision stage gap relative to Consideration (-1.06 points in ChatGPT, -1.14 points in Gemini) signals a content weakness. When buyers are at the decision stage, they ask for specific validation: certifications, contract references, client case studies, outcome data, SLA terms, implementation timelines. The AI answers these queries with whatever specific, verifiable content it has indexed.

Veolia's Decision-stage content in AI training data appears thinner than its Consideration-stage presence. The brand is recognized and positioned well when compared broadly, but the AI struggles to provide dense, specific, actionable Decision-stage details — so it either positions Veolia further down the response list or provides less detail per mention, depressing both the position score and the BIS.

The sentiment drop at Decision stage is also meaningful: ChatGPT Decision sentiment (+0.364) is the lowest of the three stages — 0.028 points below Consideration (+0.392) and 0.020 points below Awareness (+0.384). At Decision stage, queries about contracts, pricing, or compliance generate more cautious, hedged AI language, which pulls sentiment down slightly.

---

14.5 Awareness: Volume Without Quality in Gemini

Awareness shows an interesting engine divergence:

Stage	ChatGPT BIS	Gemini BIS	ChatGPT Responses	Gemini Responses
Awareness	47.48	46.19	737	999
Consideration	49.78	48.50	684	909
Decision	48.72	47.36	645	828

In ChatGPT, Awareness generates the fewest mentions (950) but also the lowest BIS (47.48) — the volume-quality relationship is consistent. In Gemini, Awareness generates the most responses (999) but also the lowest BIS (46.19) — Gemini is generating more Veolia mentions at Awareness, but with weaker positioning.

This Gemini pattern — high awareness volume, low awareness quality — suggests that Gemini surfaces Veolia in broad discovery queries but places it less prominently than in comparison queries. Awareness responses in Gemini tend to be longer lists (8–12 companies) where Veolia appears but not necessarily first or second.

Gemini's position score at Awareness (0.191) is notably lower than at Consideration (0.234) — a gap of 0.043 points, the largest inter-stage positional gap in the Gemini data. This confirms that Gemini's Awareness responses are generating more mentions but with worse placement.

---

14.6 Engine Consistency: Do ChatGPT and Gemini Agree?

Both engines agree on the funnel ranking: Consideration > Decision > Awareness. This consistency across independent AI systems strengthens the strategic signal — the pattern is structural, not engine-specific.

Stage	ChatGPT Rank	Gemini Rank	Agreement
Consideration	1st	1st	✓
Decision	2nd	2nd	✓
Awareness	3rd	3rd	✓

The absolute BIS gap between ChatGPT and Gemini is consistent across stages: ChatGPT outperforms Gemini by approximately 1.3–1.5 points at each funnel stage. This is consistent with the global BIS gap (48.6 CG vs 47.3 GM) and does not indicate a funnel-specific engine anomaly.

The one area of divergence is sentiment at Awareness: Gemini's Awareness sentiment (+0.421) is higher than ChatGPT's (+0.384), and also higher than Gemini's own Consideration sentiment (+0.410). Gemini generates more positive framing of Veolia in discovery-stage responses than ChatGPT does — but this positive sentiment doesn't translate into better BIS because Gemini's Awareness position scores remain the weakest across all funnel stages.

---

14.7 Strategic Implication

Stage	Situation	Priority	Action
Consideration	BIS 49.78 CG / 48.50 GM — strong	Maintain	Ensure comparison-ready content: structured comparisons, technical specs, multi-domain coverage
Decision	BIS 48.72 CG / 47.36 GM — gap vs Consideration	Biggest opportunity	Publish certifications, outcome data, client references, case studies with measurable results — content the AI can cite as specific validation
Awareness	BIS 47.48 CG / 46.19 GM — weakest	Secondary	Category-level thought leadership: industry reports, statistics-heavy content, authoritative positioning on water/waste/energy trends

The biggest single opportunity in the funnel analysis is Decision-stage content depth. Veolia is already winning the Consideration battle — when buyers are comparing, the brand shows up. The failure point is the final step: when a buyer asks the AI to validate Veolia specifically, the AI doesn't have enough specific, citable material to answer with confidence. A library of outcome-focused case studies, certification listings, and project-specific data — published in AI-indexable formats — would close the 2.30-point gap between Consideration and Awareness BIS and push Decision-stage performance above 50.0.

The decision gap is not a brand awareness problem. It is a content depth problem.

Page 15 — What the AI Says: Verbatims

---

15.1 Selection Criteria

The verbatims below are drawn directly from the ChatGPT and Gemini response CSV exports (verbatims_positive, verbatims_neutral, verbatims_negative). All fragments are exact quotes from AI responses. Only verbatims where "Veolia" is explicitly named have been included. Selection covers all four markets, all three product lines, and both engines, prioritising representative range over statistical frequency.

---

15.2 Positive Verbatims — The AI Recommends Veolia

🏭 Hazardous Waste / Spain — Veolia Recommended First

"Veolia stands out among the companies you've mentioned—Veolia, Tradebe, Befesa, and FCC Medio Ambiente—when evaluating reliability in handling hazardous industrial waste in ports, particularly concerning turnaround time, traceability, and permitted treatment capacity. Veolia has significantly expanded its hazardous waste treatment capabilities. In June 2025, Veolia announced plans to add 530,000 tonnes of new hazardous waste annual treatment capacity."

Engine: ChatGPT | Persona: Port Energy Infrastructure Manager - Spain | Sentiment: 5/5 | Position: 0 (first mention)

Why it works: The AI doesn't hedge — it picks Veolia as the standout option over three named competitors, and backs the recommendation with a specific 2025 capacity expansion figure (530,000 tonnes). This is the closest Veolia gets to a direct AI endorsement in the dataset.

---

💧 Water / United States — NRW Reduction with Measurable Outcomes

"Veolia has demonstrated significant success in non-revenue water (NRW) reduction programs in the United States, often with guaranteed performance and utility references. For instance, Veolia's Regulated Water division saved 1.5 billion gallons of non-revenue water over five years through advanced metering systems and leak detection technology, achieving an 84% network efficiency, surpassing its 80% target. This resulted in substantial financial benefits, including $1.1 million in savings."

Engine: ChatGPT | Persona: Urban Energy Planning Advisor - United States | Sentiment: 5/5 | Position: 0 (first mention)

Why it works: Specific, citable metrics (1.5 billion gallons, 84% network efficiency, $1.1 million savings) drive the AI's confidence. This is exactly the kind of outcome-focused content that generates high BIS: the AI cites the data, not just the company name.

---

🏆 Water / United States — "Top Environmental Firm" Attribution

"Veolia North America: Recognized as the top environmental firm in the U.S. for five consecutive years, Veolia offers comprehensive wastewater treatment solutions that focus on sustainability and energy optimization."

Engine: ChatGPT | Persona: Urban Energy Planning Advisor - United States | Sentiment: 5/5 | Position: 0 (first mention) | Source cited: veolianorthamerica.com

Why it works: The AI attributes an external accolade ("top environmental firm for five consecutive years") and links directly to a Veolia-owned URL. This is narrative control at its best: the source is veolianorthamerica.com, the claim is positive, and the positioning is first.

---

⚡ Bioenergy / Spain — Strongest for District Heating Integration

"Veolia appears to be the strongest contender for integrating waste-to-energy (WtE) heat off-take with district heating networks in Spanish cities. Veolia has a clear and documented history of developing and operating biomass district heating systems in Spain, such as the one in Móstoles. They have explicit district heating integration capabilities documented."

Engine: Gemini | Persona: Regional Waste & Resource Recovery Director - Spain | Sentiment: 5/5 | Position: 0 (first mention)

Why it works: Gemini picks Veolia over SUEZ, Urbaser, and FCC Medio Ambiente with a specific Spanish project reference (Móstoles). Local project documentation directly translates into AI recommendation positioning.

---

🌍 Hazardous Waste / Middle East — Vision 2030 Contract Authority

"Veolia generally demonstrates a stronger and more established track record for long-term hazardous waste treatment contracts tied to municipal KPIs and national diversion targets like Saudi Vision 2030. Veolia is a global leader in hazardous waste management, with a significant presence in the UAE and the wider GCC region."

Engine: Gemini | Persona: Municipal Circular Economy Manager - Middle East | Sentiment: 5/5 | Position: 28 (in a longer list, still positive framing)

Why it works: Gemini frames Veolia as the default authority for Vision 2030-aligned contracts — a high-intent, high-value procurement context. The "stronger and more established track record" phrasing shows the AI is making a comparative judgment in Veolia's favour.

---

💧 Water / Spain — Ecological Transformation with Investment Data

"Veolia has demonstrated a strong commitment to ecological transformation in Spain, investing over €360 million between 2022 and 2023. Their approach integrates advanced pressure management, smart metering, and AI-based predictive analytics to identify leaks and reduce non-revenue water (NRW) proactively."

Engine: ChatGPT | Persona: Urban Energy Planning Advisor - Spain | Sentiment: 4/5 | Position: 0 (first mention)

Why it works: €360 million investment figure + named technologies (smart metering, AI-based predictive analytics) + local market anchor (Spain). Three GEO signals in one response: data, technical specificity, and geographic relevance.

---

15.3 Neutral Verbatims — Information Without Recommendation

⚗️ Water Technologies / Middle East — Capability Listed, Not Prioritised

"Veolia has demonstrated significant expertise in high-recovery reverse osmosis (RO) and polishing technologies for boiler feed water reuse in the Gulf Cooperation Council (GCC) countries, particularly in Saudi Arabia and the UAE. Their involvement in large-scale projects, such as the Sadara SWRO plant in Saudi Arabia and the Fujairah 2 RO desalination plant in the UAE, underscores their capability in this domain."

Engine: ChatGPT | Persona: Industrial Water Efficiency Manager - Middle East | Sentiment: 3/5 | Position: 2

Pattern: The AI confirms Veolia's capability and cites real projects (Sadara, Fujairah 2), but does not select it as the strongest option. The response lists Veolia alongside SUEZ and regional EPCs without giving a clear winner. This is the "competent but not recommended" zone — high capability recognition, low directional signal.

---

🧪 ZLD / Spain — Technical Credibility, No Differentiation

"Veolia offers comprehensive solutions for mining wastewater treatment, encompassing biological treatment, thermal evaporators, and crystallizers. Their approach emphasizes decarbonization, depollution, and water reuse, aligning with global sustainability goals."

Engine: ChatGPT | Persona: Industrial Water Efficiency Manager - Spain | Sentiment: 3/5 | Position: 6

Pattern: Veolia is included as one of three options (alongside SUEZ and Acciona) with no differentiation. The description is accurate but generic — no outcome data, no specific project references, no concrete differentiator. The AI knows what Veolia does but doesn't know why it's better.

---

15.4 Critical Verbatims — Compliance Framing and Competitive Displacement

⚠️ Hazardous Waste / Australia — Compliance Violation Record

"Compliance Track Record: Veolia has faced several environmental violations in Australia, including waste and water pollution violations in 2021 and 2024, with fines totaling approximately $50,000."

Engine: ChatGPT | Persona: Municipal Water Infrastructure Manager - Australia | Sentiment: 2/5 | Position: 0 (mentioned first — negatively)

Pattern: The AI opens the compliance section with Veolia's violation record. The source cited is a third-party violation tracking database — not a Veolia-owned property. This is the clearest example of narrative risk in the dataset: the AI is citing adverse information from an external source that Veolia does not control. The $50,000 fine figure appears small, but the framing ("faced several environmental violations in 2021 and 2024") creates a disproportionate negative impression at the Decision stage.

---

📊 Digital Water / Middle East — SUEZ Leads on Platform Detail

"Both SUEZ and Veolia have extensive experience integrating their digital water platforms with existing SCADA and GIS systems. SUEZ, through its AQUADVANCED® platform, is designed to integrate with SCADA and GIS for real-time management. Veolia offers comparable integration, but SUEZ's platform documentation is more detailed on cybersecurity posture and licensing transparency."

Engine: Gemini | Persona: Municipal Water Infrastructure Manager - Middle East | Sentiment: 2/5 | Position: 15

Pattern: When the AI compares digital platforms, SUEZ's named product (AQUADVANCED®) gives it a specificity advantage that Veolia lacks. Veolia is "comparable" but not "leading." Named product documentation beats unnamed capabilities — a clear GEO content signal.

---

15.5 Patterns: What Does the AI Say About Veolia?

Context of Mention

Context	Engine Tendency	AI Frame
Hazardous waste — Europe/Spain	ChatGPT strong, Gemini strong	"Global leader," "Most reliable"
Water treatment — US	ChatGPT strong	"Top environmental firm," "Guaranteed performance"
Bioenergy/district heating — Spain	Gemini strong	"Strongest contender," specific project references
Water reuse — Middle East	Both engines	"Significant expertise," specific projects cited (Sadara, Fujairah 2)
Compliance record — Australia	ChatGPT critical	"Faced several violations" (violation tracker source)
Digital platforms	Gemini neutral	"Comparable" to SUEZ, less product-specific

What Triggers Strong Positive Mentions

1. Specific outcome data — "1.5 billion gallons saved," "84% efficiency," "€360 million investment" generate immediate positive framing
2. Named projects — Sadara SWRO, Fujairah 2, Móstoles biomass are cited as proof points when the AI has the data
3. Comparative queries — When prompted to choose between named competitors, the AI tends to pick Veolia first (especially in Spain hazardous waste and Middle East)
4. External accolades — "Top environmental firm for five consecutive years" (sourced from veolianorthamerica.com) transfers directly into positive AI framing

What Triggers Neutral or Critical Mentions

1. Absence of outcome data — Generic capability descriptions without metrics produce 3/5 sentiment
2. Compliance-focused queries — Any query asking about "compliance track record" or "violations" surfaces the Australia fine data
3. Named competitor products — When SUEZ has AQUADVANCED® or Cleanaway has "Hippo Station," Veolia's unnamed equivalent loses the comparison
4. Local competitor saturation — In Australia, Cleanaway's volume of local content edges Veolia in several response contexts

Page 16 — Sources Cited by the AI

---

16.1 How the AI Builds Its Veolia Narrative

Every time an AI engine mentions Veolia, it draws on a source ecosystem — the documents, pages, and references that shaped its training data and real-time retrieval. This section maps the top domains cited in responses where Veolia is explicitly mentioned (source: sources_by_target_brand.csv, both engines).

Understanding which domains dominate this ecosystem is not an academic exercise. It determines what the AI knows, how it frames it, and whether Veolia controls the story or someone else does.

Data scope: - ChatGPT: 10,619 total domain citations across 2,461 unique domains (Veolia-mention responses only) - Gemini: 38,251 total domain citations across 8,548 unique domains (Veolia-mention responses only)

---

16.2 Top 20 Domains — Cross-Engine Citation Table

#	Domain	CG Citations	GM Citations	Source Type	Notes
1	en.wikipedia.org	1,672	~5,000+	Encyclopedic	#1 by far; Veolia Wikipedia page cited 295× in CG alone
2	www.anz.veolia.com	144	711	Official (Veolia)	Veolia Australia/NZ — GM favourite
3	www.veolia.com	242	576	Official (Veolia)	Global corporate site
4	smartwatermagazine.com	62	430	Trade media	Key earned media channel; Gemini cites heavily
5	www.suez.com	136	406	Competitor	Former SUEZ — narrative competitor
6	www.xylem.com	83	379	Competitor	Water tech competitor
7	www.fluencecorp.com	80	351	Third-party tech	Decentralised water treatment
8	www.veolianorthamerica.com	200	269	Official (Veolia)	Veolia North America
9	www.usdanalytics.com	219	310	Market research	Heavy CG usage; mining water focus
10	www.cleanaway.com.au	34	309	Competitor	Veolia's main Australia rival
11	www.near-middle-east.veolia.com	108	308	Official (Veolia)	Veolia Middle East
12	www.epa.gov	64	270	Government/Regulator	US regulator; compliance context
13	www.researchgate.net	—	241	Academic	GM-heavy; peer-reviewed citations
14	www.researchandmarkets.com	38	223	Market research	Industry report aggregator
15	www.cleanharbors.com	164	195	Competitor	US hazardous waste leader
16	www.mdpi.com	—	294	Academic	Open-access journals; Gemini-indexed
17	www.mordorintelligence.com	49	171	Market research	Industry analysis
18	www.watertechnologies.com	68	164	Third-party tech	Water tech portal
19	tma.es	61	332	Competitor (Spain)	Spanish hazardous waste specialist
20	www.kenresearch.com	122	69	Market research	Spain hazardous waste market focus

---

16.3 Who Controls Veolia's AI Narrative?

ChatGPT — Source Type Distribution

Source Type	Key Domains	Total Citations	% of CG Total
Encyclopedic	en.wikipedia.org, es.wikipedia.org	~1,729	16.3%
Official (Veolia)	veolia.com, veolianorthamerica.com, anz.veolia.com, near-middle-east.veolia.com, veoliawatertech.com, veoliawatertechnologies.com, hazardouswasteeurope.veolia.com (+others)	~956	9.0%
Market Research	usdanalytics.com, kenresearch.com, grandviewresearch.com, pmarketresearch.com, mordorintelligence.com, dataintelo.com	~617	5.8%
Competitors	cleanharbors.com, suez.com, xylem.com, cleanaway.com.au, tma.es, acciona.com, remondis-australia.com.au	~521	4.9%
Trade Media	smartwatermagazine.com, businesswire.com, waterworld.com, globenewswire.com, cincodias.elpais.com	~294	2.8%
Government	epa.gov, dcceew.gov.au	~93	0.9%
All other	2,400+ long-tail domains	~6,409	60.4%

Gemini — Source Type Distribution

Source Type	Key Domains	Total Citations	% of GM Total
Encyclopedic	en.wikipedia.org (global), es.wikipedia.org	~5,200+	~13.6%
Official (Veolia)	anz.veolia.com, veolia.com, near-middle-east.veolia.com, veolianorthamerica.com, veoliawatertechnologies.com (+others)	~2,842	7.4%
Trade Media	smartwatermagazine.com, businesswire.com, waterworld.com	~700+	~1.8%
Competitors	suez.com, cleanharbors.com, cleanaway.com.au, tma.es, xylem.com, acciona.com	~1,600+	~4.2%
Market Research	researchandmarkets.com, mordorintelligence.com, kenresearch.com, pmarketresearch.com	~800+	~2.1%
Academic	researchgate.net, mdpi.com	~535	1.4%
All other	8,000+ long-tail domains	~27,400	~71.6%

---

16.4 Problematic Sources — Narrative Risk

Competitor Domains in Veolia Responses

Domain	CG Citations	GM Citations	Why It's a Risk
www.cleanharbors.com	164	195	The AI cites Clean Harbors content when answering US hazardous waste queries — sometimes instead of Veolia content
www.suez.com	136	406	Post-merger, SUEZ still appears as a separate reference in AI responses — diluting Veolia's position as the global water leader
www.cleanaway.com.au	34	309	Gemini heavily weights Cleanaway for Australian hazardous waste queries; it competes directly with Veolia ANZ
tma.es	61	332	Spanish hazardous waste specialist consistently cited alongside Veolia in Spain-focused responses
www.xylem.com	83	379	Water technology competitor frequently co-cited with Veolia in water treatment comparisons

The SUEZ effect: Despite the Veolia–SUEZ merger (Veolia acquired SUEZ in 2022), the AI treats them as separate entities. suez.com still generates 136 (CG) and 406 (GM) citations in Veolia-mention responses. This means the AI is citing a competitor brand that is now part of Veolia — a narrative integration gap.

Compliance and Violation Sources

The verbatim analysis (page 15) identified that in Australia, the AI cites violation tracking databases (e.g., violationtracker.com) when asked about compliance records. This source type does not appear in the top 20 domains but is cited at the Decision stage — exactly when buyers are making final judgements.

---

16.5 Top Individual URLs Most Cited in Veolia Responses

ChatGPT — Top URLs

URL	Citations	Type
en.wikipedia.org/wiki/Veolia_Environmental_Services	295	Encyclopedic (Veolia)
en.wikipedia.org/wiki/Clean_Harbors	389	Encyclopedic (Competitor)
en.wikipedia.org/wiki/Tradebe	361	Encyclopedic (Competitor)
kenresearch.com/spain-hazardous-waste-management-market	98	Market research
austransgroup.com.au/services/hazardous-waste/	95	Third-party competitor (Australia)
tma.es/en/hazardous-waste/	73	Competitor (Spain)

Gemini — Top URLs (Veolia-related)

URL	Citations	Type
smartwatermagazine.com/news/veolia/decarbonizing-gccs-water-sector...	106	Trade media (Veolia coverage)
cleanharbors.com/services/technical-services/waste-disposal-services	252	Competitor
cleanearthinc.com/services/waste-disposal-transportation	252	Competitor
indaver.com/locations/spain	251	Competitor (Spain)
cleanaway.com.au/hazardous-waste	184	Competitor (Australia)
austransgroup.com.au/hazardous-waste/	149	Third-party

Key insight: In ChatGPT, the most-cited Veolia-specific URL is the Wikipedia page for "Veolia Environmental Services" (295 citations). In Gemini, the most-cited Veolia-specific URL is a Smart Water Magazine article about GCC decarbonisation (106 citations). Veolia's own corporate URLs — while present — do not dominate the individual URL ranking.

---

16.6 Narrative Control Summary

The AI's Veolia story is primarily written by three parties:

1. Wikipedia — The single largest source by volume (1,672 CG citations). Veolia does not control this content directly. The Veolia Environmental Services Wikipedia article (295 CG citations alone) is the AI's primary reference document for foundational Veolia facts.

2. Veolia's own digital ecosystem — 956 citations (CG) / 2,842 citations (GM) across all Veolia domains. Significant but secondary to Wikipedia. Represents 9.0% (CG) and 7.4% (GM) of the total citation footprint.

3. Market research and trade media — usdanalytics.com, kenresearch.com, smartwatermagazine.com collectively shape the technical and competitive framing. These sources are not hostile, but they are not controlled by Veolia either.

The combined share of the top 3 source types in ChatGPT: - Wikipedia (16.3%) + Veolia owned (9.0%) + Market research (5.8%) = 31.1% of total citations - The remaining 68.9% comes from long-tail domains, competitors, trade media, and government sources that Veolia cannot directly influence.

Page 17 — Owned Media: GEO Assessment

---

17.1 Veolia's Official Domain Ecosystem

Veolia operates a large, geographically distributed web infrastructure. The official_sources.csv files enumerate every Veolia-owned domain that the AI cited when mentioning Veolia. The picture is more complex than a single corporate site: Veolia has regional subdomains, water technology subbrands, hazardous waste portals, and market-specific URLs — all of which the AI indexes and cites independently.

---

17.2 Full Owned Media Inventory — Citation Data

ChatGPT — Official Veolia Sources

Domain	Citations	Unique URLs	Content Type
www.veolia.com	243	79	Global corporate
www.veolianorthamerica.com	200	43	North America operations
www.anz.veolia.com	144	55	Australia / New Zealand
www.near-middle-east.veolia.com	108	22	Middle East operations
www.veoliawatertech.com	46	25	Water technology brand
www.veoliawatertechnologies.com	42	22	Water technologies global
www.anz.veoliawatertechnologies.com	26	8	Water tech ANZ
www.hazardouswasteeurope.veolia.com	21	9	European hazardous waste
info.veolianorthamerica.com	13	8	North America info/lead gen
www.middle-east.veoliawatertechnologies.com	11	8	Middle East water tech
www.southeastasia.veolia.com	10	7	SE Asia operations
www.veolia.es	8	7	Spain local
www.veolia.bg	8	5	Bulgaria (minor)
www.latinoamerica.veolia.com	7	5	Latin America
www.vigie-sa.veolia.com	7	3	Safety/inspection product
www.veolia.in	6	3	India
www.veoliawatertechnologies.es	5	5	Spain water tech
services.veolia.com.au	5	3	Australia services
www.remediation.veolia.com	5	3	Remediation vertical
www.africa.veolia.com	5	3	Africa operations
blog.veolianorthamerica.com	4	4	NA blog
Other domains (19)	~13	~18	Misc. markets
Total ChatGPT	~956	~338

Gemini — Official Veolia Sources

Domain	Citations	Unique URLs	Content Type
www.anz.veolia.com	724	123	Australia / New Zealand
www.veolia.com	582	111	Global corporate
www.near-middle-east.veolia.com	314	38	Middle East operations
www.veolianorthamerica.com	276	74	North America operations
www.veoliawatertechnologies.com	156	51	Water technologies global
www.veoliawatertech.com	146	46	Water technology brand
www.anz.veoliawatertechnologies.com	117	22	Water tech ANZ
www.middle-east.veoliawatertechnologies.com	63	14	Middle East water tech
www.hazardouswasteeurope.veolia.com	62	17	European hazardous waste
www.veolia.co.uk	44	22	United Kingdom
www.veolia.es	43	20	Spain local
info.veolianorthamerica.com	39	11	North America info
www.engineering-consulting.veolia.com	29	5	Engineering consulting
blog.veoliawatertechnologies.com	27	4	Water tech blog
blog.veolianorthamerica.com	24	8	NA blog
www.asia.veoliawatertechnologies.com	23	11	Asia water tech
www.latinoamerica.veolia.com	21	8	Latin America
Other domains (34)	~251	~120	Misc. blogs, markets
Total Gemini	~2,842	~705

---

17.3 Owned vs Earned Media — The Core Ratio

The central GEO question: what proportion of AI citations come from sources Veolia controls vs. sources it influences but does not own?

Engine	Total Domain Citations	Veolia Owned Citations	% Owned	% Third-Party
ChatGPT	10,619	956	9.0%	91.0%
Gemini	38,251	2,842	7.4%	92.6%

Veolia owns approximately 9% of its AI citation footprint in ChatGPT and 7.4% in Gemini. The remaining 91–93% of sources are third parties: Wikipedia, market research firms, trade publications, competitor sites, government regulators, and long-tail industry portals.

This is the foundational GEO challenge. Even if every Veolia-owned page were perfectly optimised, only 9% of the narrative input would improve. The remaining 91% requires earned media strategy, Wikipedia stewardship, and third-party content influence.

---

17.4 Content Types Most Cited from veolia.com

Across 79 unique URLs from www.veolia.com (ChatGPT) and 111 (Gemini), the AI's preferred content types are:

Content Category	Engine Preference	Representative URL Pattern	Why the AI Cites It
Press releases / newsroom	ChatGPT, Gemini	/newsroom/, /media/, /press-releases/	Factual, dated, specific project data
Industry reports / white papers	Gemini	/en/our-media/publications/	Dense data; authoritative framing
Regional operations pages	Both (regional subdomains)	anz.veolia.com, near-middle-east.veolia.com	Local context + specific contracts
Product/service pages (Water Tech)	Both	veoliawatertechnologies.com, veoliawatertech.com	Technical specs, technology descriptions
Hazardous waste portals	ChatGPT	hazardouswasteeurope.veolia.com	Regulatory alignment, capacity data

Key observation: Veolia's water technology subdomains (veoliawatertechnologies.com, veoliawatertech.com, anz.veoliawatertechnologies.com) generate 114 CG citations and 419 GM citations combined — nearly as much as the global corporate site. The AI treats Veolia's technology-focused properties as distinct reference sources from the corporate brand. This fragmentation is a double-edged sword: more indexed surfaces, but divided citation weight.

---

17.5 Regional Owned Media Performance

Region	Primary Official Domain	CG Citations	GM Citations	Performance Assessment
Global	veolia.com	243	582	Strong; corporate anchor
North America	veolianorthamerica.com	200	276	Strong; best outcome-data content
Australia/NZ	anz.veolia.com	144	724	Gemini dominant; local operations content
Middle East	near-middle-east.veolia.com	108	314	Both engines; Gulf projects drive citations
Spain	veolia.es	8	43	Underperforming for market size
Europe Hazwaste	hazardouswasteeurope.veolia.com	21	62	Niche but functional
UK	veolia.co.uk	3	44	Strong Gemini, weak ChatGPT

The Spain gap is the clearest owned media problem. veolia.es generates only 8 citations in ChatGPT and 43 in Gemini — despite Spain being the 3rd-largest market in this study and a key Hazardous Waste and Water market. By comparison, anz.veolia.com (a similar regional domain) generates 144 CG / 724 GM citations. This suggests that veolia.es has significantly less AI-indexable content than its Australian or North American equivalents.

---

17.6 Owned Media Efficiency — Citations per URL

Domain	Engine	Citations	Unique URLs	Citations/URL	Assessment
www.veolia.com	CG	243	79	3.1	Moderate efficiency
www.veolianorthamerica.com	CG	200	43	4.7	High efficiency
www.anz.veolia.com	GM	724	123	5.9	High efficiency
www.near-middle-east.veolia.com	GM	314	38	8.3	Very high efficiency
www.hazardouswasteeurope.veolia.com	GM	62	17	3.6	Moderate
www.veolia.es	GM	43	20	2.2	Low — content too generic
www.engineering-consulting.veolia.com	GM	29	5	5.8	High efficiency, low volume

near-middle-east.veolia.com has the highest citation-per-URL ratio (8.3 in Gemini): each indexed page generates 8+ citations on average. This mirrors the La Caixa Observatori Social pattern — a focused domain with dense, factual content generates outsized GEO returns per page published. Replicating this pattern (data-rich, project-specific content) at veolia.es would be the highest-ROI owned media move in Spain.

---

17.7 GEO Benchmark: What Does 9% Owned Media Mean?

For context, the 9% owned media share in ChatGPT means:

• For every 100 times ChatGPT cites a source when mentioning Veolia, 9 are Veolia-owned pages
• Wikipedia alone contributes 16.3% — nearly twice the owned media share
• Competitor sites collectively contribute ~4.9% — meaning competitors are writing nearly half as much of Veolia's narrative as Veolia itself

Industry reference: A brand with strong GEO optimisation typically achieves 15–25% owned media share in AI citations, per observed benchmarks in B2B sectors. At 9% (CG) and 7.4% (GM), Veolia is below this range — but the gap is closeable. The ceiling for owned media is not 100%: Wikipedia and third-party market reports will always dominate in volume. The realistic GEO target is 15–18% owned media share — achievable by doubling indexable content depth on regional sites and improving the Spain domain specifically.

The most actionable lever: veolianorthamerica.com achieves 4.7 citations/URL with 43 URLs indexed. If veolia.es replicated this efficiency (currently 2.2 citations/URL), and published comparable content depth (target: 80+ indexed URLs), it could generate 370+ ChatGPT citations — versus the current 8. That single domain improvement would shift the Spain market BIS meaningfully.

Page 18 — Narrative Control Strategy

---

18.1 Who Writes the Veolia Story in AI?

The source analysis from pages 16 and 17 establishes a clear picture: Veolia does not control its own AI narrative. That statement is not a criticism — no brand in a competitive B2B sector controls the majority of its AI citation footprint. But the specific distribution matters enormously for strategy.

The entities that shape what ChatGPT and Gemini say about Veolia, in order of citation volume:

Source Category	ChatGPT Share	Gemini Share	Control Level
Encyclopedic (Wikipedia)	16.3%	~13.6%	🔴 Low — community-edited
Veolia owned media	9.0%	7.4%	✅ High — full control
Market research	5.8%	~2.1%	🟡 Low — licensable, not editable
Competitors	4.9%	~4.2%	🔴 Zero — adversarial
Trade media	2.8%	~1.8%	🟡 Medium — via PR and earned coverage
Government / Regulator	0.9%	0.7%	🔴 Zero — compliance-driven
Long-tail / Other	60.4%	~70.2%	🔴 Near-zero — diffuse

The working conclusion: Veolia fully controls less than 10% of its AI citation inputs. The remaining 90%+ is shaped by forces that require influence, not ownership: Wikipedia stewardship, earned media strategy, market research seeding, and competitor co-occurrence management.

---

18.2 Source Type Dominance Analysis

Wikipedia: The Uncontrolled Foundation

Wikipedia is the single most-cited domain in ChatGPT Veolia responses (1,672 citations — 16.3% of total). More specifically, the Wikipedia page for "Veolia Environmental Services" is the 3rd most-cited individual URL in the entire ChatGPT dataset, at 295 citations.

This means the Wikipedia article about Veolia is the AI's primary reference document for foundational facts — company history, size, controversies, global footprint. Veolia does not control this content. The article is community-maintained and, critically, reflects whatever is documented there — including any historical controversies, regulatory fines, or negative media coverage.

The Australia compliance incident that appears in verbatims (environmental violations in 2021 and 2024) likely derives partly from Wikipedia-indexed content and violation-tracking databases. This is not a Wikipedia article problem per se — it's a signal that third-party sources of negative content exist and are AI-indexable.

Market Research: The Framing Layer

usdanalytics.com (219 CG citations), kenresearch.com (122 CG), grandviewresearch.com (76 CG), and pmarketresearch.com (62 CG) collectively contribute ~617 ChatGPT citations — more than half of Veolia's entire owned media footprint. These platforms publish market reports that the AI treats as authoritative industry framing.

These reports typically describe Veolia as a "key player" or "prominent vendor" within market landscapes — neutral but not differentiating language. When the AI cites a market report, it produces the 3/5 sentiment verbatims: technically accurate, but generic.

The market research problem: Veolia's messaging in these reports is mediated by analysts who are not brand advocates. The AI inherits this analyst framing directly. The strategic implication is to provide market research firms with specific, outcome-focused data that they incorporate into future reports — indirectly improving the AI's narrative tone.

Trade Media: The Quality Multiplier

Smart Water Magazine (62 CG / 430 GM) is the standout trade media source. One specific article — "Decarbonizing the GCC's water sector: replacing desalination with reused water" — generates 106 Gemini citations, making it the most-cited individual Veolia-specific URL in the entire Gemini dataset. This demonstrates the GEO leverage of a single well-placed, data-rich trade media article.

Compared to the effort required to publish and index a corporate page, a trade article that generates 106 citations represents exceptional earned media ROI.

Competitors: The Narrative Dilution Problem

Competitor domains generate 521 CG and ~1,600 GM citations in Veolia-mention responses. This doesn't mean the AI is citing competitors instead of Veolia — it means the AI cites them alongside Veolia, in comparative responses. The practical effect:

• When comparing Veolia and Clean Harbors, the AI often cites both domains — giving Clean Harbors equal content authority
• When suez.com appears in Veolia responses (136 CG / 406 GM), the AI is still treating SUEZ as an independent entity despite the merger
• When tma.es appears in Spanish hazardous waste queries (61 CG / 332 GM), TMA is extracting citation equity from the same response context as Veolia

---

18.3 ChatGPT vs Gemini — Source Ecosystem Divergence

The two engines use different source infrastructures, which produces different narrative structures:

Dimension	ChatGPT	Gemini
Top non-owned source	en.wikipedia.org (1,672)	vertexaisearch.cloud.google.com (1,085) + Wikipedia
Owned media citations	956 (9.0%)	2,842 (7.4%)
Competitor citations	~521	~1,600+
ANZ domain weight	anz.veolia.com: 144	anz.veolia.com: 711
Spain domain weight	veolia.es: 8	veolia.es: 43
Trade media	smartwatermagazine: 62	smartwatermagazine: 430
Academic sources	Minimal	researchgate.net: 241, mdpi.com: 294
Market research	usdanalytics.com: 219	Distributed across many

Key divergences: 1. Gemini weights Australian content far more heavily. anz.veolia.com gets 711 Gemini citations vs 144 ChatGPT — a 5:1 ratio. This amplifies Veolia's Australian positioning in Gemini but also means Cleanaway's local Australian content (309 GM citations) is more competitive in Gemini than in ChatGPT. 2. Gemini relies heavily on academic and scientific sources (researchgate.net 241, mdpi.com 294). ChatGPT does not meaningfully cite these. This means Gemini's technical framing of Veolia reflects academic/research discourse more than ChatGPT does — potentially more rigorous but also more likely to include critical environmental research. 3. Smart Water Magazine is a Gemini amplifier. At 430 Gemini citations vs 62 ChatGPT, this single trade publication has disproportionate influence over Veolia's Gemini narrative. One article about GCC decarbonisation generates 106 Gemini citations alone. 4. ChatGPT relies more on market research (usdanalytics.com 219 CG, kenresearch.com 122 CG). Gemini distributes this across a wider range of sources.

---

18.4 Narrative Vulnerability Points

Vulnerability 1: Wikipedia — Unmonitored Foundation

The Wikipedia "Veolia Environmental Services" article (295 CG citations) is the AI's primary Veolia reference document. Its content — including any controversies, historical regulatory issues, or incomplete information — flows directly into AI responses. Veolia does not control this article.

Risk level: High
Immediacy: This affects every single AI response that cites Wikipedia for Veolia context.

Vulnerability 2: Compliance/Violation Databases — Australia Exposure

ChatGPT cites violation tracking sources at the Decision stage in Australia. The result (verbatim, page 15): "Veolia has faced several environmental violations in Australia, including waste and water pollution violations in 2021 and 2024, with fines totaling approximately $50,000." The fine is small; the AI framing is not.

Risk level: Medium (localised to Australia, Decision stage)
Immediacy: Active now — cited in current responses.

Vulnerability 3: SUEZ — Post-Merger Brand Fragmentation

suez.com still generates 136 CG / 406 GM citations in Veolia-mention responses. The AI treats SUEZ as a peer competitor of Veolia, not as a Veolia subsidiary. This means: - The AI sometimes recommends SUEZ instead of Veolia in European water contexts - SUEZ's citation volume dilutes Veolia's competitive positioning in AI responses - The narrative of "Veolia absorbed the best of SUEZ" is not reaching the AI's training data

Risk level: High (structural, affects competitive scoring)
Immediacy: Ongoing — requires long-term content integration strategy.

Vulnerability 4: Spain — Owned Media Vacuum

veolia.es generates 8 ChatGPT citations (vs anz.veolia.com's 144). Spain is the 3rd-largest market in this study and a key competitive battleground for hazardous waste (Tradebe, TMA) and water (Aqualia, SUEZ). The AI's Spanish narrative is written almost entirely by third parties — market reports, Spanish media, competitor sites.

Risk level: High (affects BIS in a priority market)
Immediacy: Content investment would yield results within 3–6 months.

Vulnerability 5: Named Product Gap

When Gemini compares digital water platforms, SUEZ's "AQUADVANCED®" and Cleanaway's "Hippo Station" give those brands a specificity advantage. Veolia offers comparable platforms but they are not named prominently enough in AI-indexed content to match. This is a Consideration and Decision-stage gap.

Risk level: Medium
Immediacy: Addressable through product-naming content strategy.

---

18.5 Actionable Recommendations — Based on Source Data

Priority 1 — Wikipedia Stewardship (Highest Urgency)

Action: Commission a professional Wikipedia audit of the "Veolia" and "Veolia Environmental Services" articles in English, Spanish, French, and German. - Verify that the merger with SUEZ is accurately reflected and SUEZ is positioned as a Veolia subsidiary - Ensure all quantitative impact data (water treated, countries served, employees, revenues) is current and sourced to verifiable references - Add GCC/Middle East and Australia project references that are currently underrepresented - Prepare primary sources (annual reports, press releases with specific data) that Wikipedia editors can reference — do not edit directly

Expected impact: Improved accuracy and sentiment in the 16.3% of ChatGPT citations that derive from Wikipedia.

Priority 2 — veolia.es Content Overhaul (Highest ROI)

Action: Rebuild veolia.es content depth to match veolianorthamerica.com standards. - Target: 80+ indexed URLs (currently estimated ~20) - Content format: outcome-focused case studies (with measurable results), service specification pages, Spanish regulatory compliance certifications, market-specific project references (Tradebe comparison angles, Aqualia/NRW positioning) - Technical: structured data markup (Schema.org/Organization, Schema.org/Service), FAQ sections for key Decision-stage queries, Spanish-language tone calibrated for procurement managers - Benchmark: veolianorthamerica.com achieves 4.7 citations/URL. If veolia.es reaches 3.5 citations/URL at 80 URLs = 280 ChatGPT citations vs current 8. Spain BIS impact: estimated +1.5–2.5 BIS points.

Expected impact: Spain BIS improvement from 47.40 (CG) to ~49–50.

Priority 3 — Smart Water Magazine Partnership (Gemini Amplification)

Action: Establish a recurring content partnership with Smart Water Magazine. - The single most-cited Gemini domain in Veolia responses (430 GM citations from 349 Gemini responses) - One article already generates 106 Gemini citations — the highest individual Veolia URL in Gemini - Target: 4–6 data-rich articles per year covering GCC water security, Australian water treatment, Spanish NRW, and US wastewater energy efficiency - Each article should contain: specific project data, measurable outcomes, Veolia-attributed technology names, and local market context

Expected impact: +50–150 Gemini citations per article, strengthening Gemini narrative quality across all markets.

Priority 4 — SUEZ Integration Narrative (Structural Fix)

Action: Publish a dedicated content series positioning Veolia as the inheritor and integrator of SUEZ capabilities. - Target: 3–5 long-form articles on veolia.com and veoliawatertechnologies.com: "How Veolia absorbed SUEZ expertise into X service line" - Specific angle: AQUADVANCED® platform and SUEZ water technologies are now Veolia capabilities — name them explicitly - Distribute via press releases (businesswire.com is already indexed) and trade media (waterworld.com, smartwatermagazine.com)

Expected impact: Reduction of "SUEZ vs Veolia" framing in AI responses; improvement in Consideration-stage competitive scoring.

Priority 5 — Australia Compliance Narrative Management

Action: Publish proactive transparency content on services.veolia.com.au and anz.veolia.com addressing the 2021/2024 incidents. - Format: A "Compliance and Environmental Standards" page with: incident description, resolution actions taken, current certification status, proactive monitoring protocols - Why: The AI cites violation trackers because they are the only specific, dated source. A Veolia-owned response document would compete for citation in compliance-query contexts - Supplement with Cleanaway comparison content: "Why Veolia leads Australia on environmental performance" — outcome-focused, not defensive

Expected impact: Reduction of 2/5 sentiment verbatims in Australian Decision-stage queries.

Priority 6 — Market Research Seeding

Action: Ensure Veolia outcome data appears in the key market research sources the AI cites. - usdanalytics.com (219 CG citations): Contact for inclusion in mining water treatment and ZLD market reports - kenresearch.com (122 CG): Spain hazardous waste market report — verify Veolia is accurately positioned - grandviewresearch.com (76 CG): Provide input for water treatment and bioenergy market research updates - These reports are not editable directly, but providing press releases, case study PDFs, and verifiable data to researchers improves the quality of Veolia's representation in their output

Expected impact: Sentiment upgrade in market research-driven responses from 3/5 to 4/5.

---

18.6 Narrative Control — Strategic Summary

Action	Timeline	Effort	Expected BIS Impact
Wikipedia audit (EN/ES/FR)	0–3 months	Medium	+0.5–1.0 BIS global
veolia.es content overhaul	1–6 months	High	+1.5–2.5 BIS Spain
Smart Water Magazine partnership	0–2 months	Low	+0.5–1.0 BIS Gemini global
SUEZ integration content series	2–6 months	Medium	+0.5–1.0 BIS Europe
Australia compliance page	1–3 months	Low	+0.5 BIS Australia
Market research seeding	Ongoing	Low	+0.3–0.5 BIS global

Combined potential: A coordinated execution of these six actions over 6 months could move Veolia's BIS from the current ~48 range to 51–53 in ChatGPT and 49–51 in Gemini — closing roughly half the gap between current positioning and the "dominant brand" threshold of 70+. The remaining gap requires longer-term content ecosystem development: more case studies, certifications at Decision stage, and sustained trade media presence.

---

Closing Verbatim

"Veolia North America: Recognized as the top environmental firm in the U.S. for five consecutive years, Veolia offers comprehensive wastewater treatment solutions that focus on sustainability and energy optimization."

Source: ChatGPT — URL cited: veolianorthamerica.com/media/newsroom/veolia-named-top-us-environmental-firm-fifth-year-row

This verbatim represents the ideal state: Veolia-owned content, external accolade cited by the AI, positive sentiment, first position. It works because veolianorthamerica.com published specific, award-backed content that the AI indexed and trusted. Replicating this model — owned content, specific claims, verifiable references — across all regional domains is the core GEO content strategy.

Page 19 — SWOT Analysis

---

What the Data Confirms

This SWOT is not an opinion. Every point below is anchored to a number from the audit. The study analysed 18,200 AI responses — 9,374 in ChatGPT and 8,826 in Gemini — across 36 personas, 4 markets, and 3 product lines. The findings are not ambiguous: Veolia has a dominant AI position with three concrete gaps and two structural threats to monitor.

---

Strengths

S1 — #1 by Share of Voice with a structural margin

Veolia leads all 15 defined competitors in both engines. SOV: 21.3% ChatGPT / 29.0% Gemini. The nearest rival is Clean Harbors at 9.9% in ChatGPT and Xylem at 11.2% in Gemini. That is a 2.1× advantage in ChatGPT and 2.6× in Gemini. No competitor comes close to replicating Veolia's cross-sector presence because none operates simultaneously across water, waste, and energy. Veolia qualifies for mentions across all 9 funnel × product-line combinations; most competitors are relevant in 1–2.

S2 — BIS lead over the competitor pool is decisive

Veolia's BIS: 48.63 ChatGPT / 47.31 Gemini vs. competitor pool average 41.63 ChatGPT / 37.02 Gemini. The gap is +6.99 points in ChatGPT and +10.29 points in Gemini. In Gemini, several competitors score below 40 (REMONDIS: 38.88, Waste Management: 38.40), amplifying Veolia's relative position. The brand does not merely appear more often — it appears earlier and with better framing.

S3 — Middle East: the strongest regional AI footprint

Middle East is Veolia's best market in both engines. Average BIS by market: 51.24 ChatGPT / 48.47 Gemini. Individual personas peak at 53.76 (Industrial Water Efficiency Manager – Middle East, ChatGPT). The Middle East advantage is structural: fewer local competitors, large-scale Gulf infrastructure projects (Sadara SWRO, Fujairah 2) that generate specific, citable AI references, and sentiment (+0.405 ChatGPT) driven by water security and infrastructure modernisation narratives — the highest-sentiment framing in the study.

S4 — Positive sentiment above market average at exceptional scale

Veolia's sentiment score: +0.381 ChatGPT / +0.417 Gemini. The competitor pool average: 0.342 ChatGPT / 0.361 Gemini. Veolia scores +0.039 above the pool in ChatGPT and +0.056 in Gemini. More significantly, it achieves this sentiment across 3,290 ChatGPT mentions and 4,555 Gemini mentions. ENGIE (0.4211) and Ameresco (0.4090) score higher in ChatGPT sentiment — but with 347 and 148 mentions respectively, not 3,290. High sentiment at high volume is structurally rare. Veolia is the only brand in this study that achieves it.

S5 — Competitive Score: Veolia dominates co-occurrence

Competitive Score measures how often Veolia outranks other entities mentioned in the same response. 0.777 ChatGPT / 0.843 Gemini — Veolia beats 84% of all co-occurring entities in Gemini, and 78% in ChatGPT. In Gemini, 0.843 is the highest Competitive Score among all 16 entities. When Veolia shares a response with competitors, it is positioned more favourably than them in 4 of every 5 cases.

---

Weaknesses

W1 — Owned media controls only 9% of the AI narrative

Veolia-owned domains account for 9.0% of all source citations in ChatGPT and 7.4% in Gemini. Wikipedia alone controls 16.3% of ChatGPT citations — nearly double the owned media share. The total owned media footprint: 956 citations (ChatGPT) across 338 unique URLs / 2,842 citations (Gemini) across 705 URLs. Competitor sites collectively contribute ~4.9% of ChatGPT citations in Veolia-mention responses, meaning competitors are authoring nearly half as much of Veolia's AI narrative as Veolia itself.

W2 — United States is the weakest market by a meaningful margin

US market BIS: 46.59 ChatGPT / 45.53 Gemini — the lowest of all four markets, 4.65 points below the Middle East in ChatGPT and 2.94 in Gemini. US hazardous waste and energy queries are dominated by domestic brands with deep English-language content: Clean Harbors, Waste Management Inc., Republic Services, Ameresco. The Urban Energy Planning Advisor – United States (BIS 42.53 Gemini) is the worst-performing single persona in the entire Gemini dataset.

W3 — Decision stage drops −1.06 BIS below Consideration

Funnel-stage BIS: Consideration 49.78 ChatGPT / 48.50 Gemini vs. Decision 48.72 ChatGPT / 47.36 Gemini. The gap: −1.06 ChatGPT / −1.14 Gemini. When buyers are ready to act — asking for certifications, contract references, case studies, outcome validation — the AI does not have enough specific, verifiable Veolia content to answer with confidence. Sentiment also drops: Decision-stage ChatGPT sentiment (+0.364) is the lowest of all three funnel stages. This is not a brand awareness problem. It is a content depth problem.

W4 — Bioenergy volume is insufficient relative to its quality

Bioenergy & Energy Efficiency achieves the highest position score in ChatGPT (0.293) and the highest BIS in Gemini (48.39). ChatGPT BIS: 50.08. Yet ChatGPT mentions in this product line: 707 — fewer than Water Technologies (986) and far fewer than Hazardous Waste (1,597). The efficiency is exceptional (maximum BIS per mention), but the volume cap limits the total narrative impact. Bioenergy is Veolia's highest-quality product line in AI — but barely anyone asks the AI about it relative to waste.

W5 — veolia.es generates 8 ChatGPT citations

Spain is the 3rd-largest market in this study. veolia.es: 8 ChatGPT citations across 7 unique URLs. anz.veolia.com: 144 ChatGPT citations across 55 unique URLs — an 18× citation gap with a comparable regional domain. Spain BIS: 47.40 ChatGPT (3rd of 4 markets). The Spanish AI narrative is written almost entirely by third parties: market reports, local competitors (Tradebe, TMA, Aqualia), and Spanish media. veolianorthamerica.com achieves 4.7 citations per URL with structured, outcome-focused content. veolia.es achieves 1.1 citations per URL.

---

Opportunities

O1 — Wikipedia stewardship: highest ROI single action

Wikipedia is the 3rd most-cited individual URL in ChatGPT for Veolia responses: 295 citations for "Veolia Environmental Services." The domain accounts for 1,672 ChatGPT citations (16.3% of all source citations). Veolia does not control this content. The same Wikipedia article that drives 295 positive citations also contains any historical controversy or incomplete information that flows into AI responses. A professional Wikipedia audit — updating the SUEZ merger integration, adding Gulf and Australian project references, sourcing all claims to primary documents — would improve the accuracy and tone of 16.3% of ChatGPT's Veolia narrative input. No other single action touches this much citation volume.

O2 — Bioenergy quality is ready to scale

The content quality is already there. BIS 50.08 in ChatGPT (2nd only to Water Technologies at 50.11), and BIS 48.39 in Gemini — the top product line. The AI already frames Veolia as the "strongest contender" for district heating integration (verbatim from page 15), citing specific Spanish projects (Móstoles biomass). The constraint is volume: 707 ChatGPT mentions vs 1,597 for Hazardous Waste. A content program that adds Bioenergy case studies, biomethane project data, and waste-to-energy outcome metrics would scale the volume without needing to rebuild the quality — it already exists.

O3 — veolia.es overhaul: 8 to 280+ citations in 6 months

near-middle-east.veolia.com achieves 8.3 citations per URL in Gemini — the highest citation efficiency of any Veolia-owned domain. veolianorthamerica.com achieves 4.7 in ChatGPT. If veolia.es reaches 3.5 citations per URL across 80 indexed URLs (from the current ~20), it would generate ~280 ChatGPT citations — up from 8. Spain BIS impact: estimated +1.5 to 2.5 points. The domain already exists; the content depth does not.

O4 — Awareness gap: the entry point needs strengthening

Awareness-stage BIS: 47.48 ChatGPT / 46.19 Gemini — the weakest funnel stage in both engines, 2.30 points below Consideration in ChatGPT and 2.31 in Gemini. Gemini generates the most Awareness responses (999 — more than Consideration at 909) but with the weakest positioning (position score 0.191). Buyers who do not yet know Veolia are the least well served. Category-level thought leadership — statistics-heavy industry reports, trend analysis in water scarcity or energy efficiency, authoritative environmental market overviews — would improve Awareness positioning without requiring prospect-specific content.

O5 — Smart Water Magazine: one article = 106 Gemini citations

Smart Water Magazine is the #1 trade media source in Gemini: 430 citations. A single article — "Decarbonizing the GCC's water sector" — generates 106 Gemini citations individually, making it the most-cited Veolia-specific URL in the entire Gemini dataset. The leverage is exceptional: one well-placed, data-rich article in this single publication generates more Gemini citations than veolia.es generates in its entire ChatGPT presence. A systematic partnership producing 4–6 articles per year would have outsized Gemini impact.

---

Threats

T1 — ENGIE leads Veolia in ChatGPT BIS

ENGIE BIS 49.18 vs. Veolia 48.63 in ChatGPT — a 0.55-point gap, driven by ENGIE's stronger sentiment (0.4211 vs. 0.3806), position score (0.2803 vs. 0.2689), and mention depth (0.2395 vs. 0.2256). Veolia only outperforms ENGIE on Competitive Score (0.7768 vs. 0.7414). ENGIE has 347 ChatGPT mentions — a fraction of Veolia's 3,290 — but each ENGIE mention in an energy context is more enthusiastic. The AI treats them as co-leaders in bioenergy queries (verified in page 15 verbatims). This is the only instance in the study where a competitor leads Veolia in BIS.

T2 — SUEZ is Veolia's permanent co-shadow

SUEZ co-occurs with Veolia in 480 ChatGPT / 630 Gemini responses — the highest co-occurrence of any entity in both engines. The AI continues to treat SUEZ as an independent competitor despite the 2022 acquisition. suez.com generates 136 ChatGPT / 406 Gemini citations in Veolia-mention responses. Every buyer reading an AI response about Veolia also reads about SUEZ in the same response. The narrative opportunity Veolia has not captured: the merger absorbed SUEZ's technology portfolio, including AQUADVANCED® — but the AI does not know this.

T3 — Wikipedia uncontrolled, compliance content indexed

The Wikipedia "Veolia Environmental Services" article (295 ChatGPT citations) is community-edited. Any negative content documented there — historical controversies, regulatory incidents, incomplete merger information — flows directly into AI responses. The Australia compliance verbatim (page 15) illustrates the active risk: "Veolia has faced several environmental violations in Australia, including waste and water pollution violations in 2021 and 2024, with fines totaling approximately $50,000." The fine is small; the AI framing is not. Decision-stage queries in Australia surface this content because it is the only specific, dated compliance source the AI has indexed.

T4 — US domestic competition is the most crowded AI landscape

The US market is the most competitive in the study. ChatGPT US responses populate with Clean Harbors (SOV 9.92%), Republic Services, Waste Management Inc., Ameresco, and ENGIE North America — all with deep English-language content volumes. Veolia's US BIS (46.59) is 4.65 points below its Middle East BIS. The Urban Energy Planning Advisor – United States (BIS 45.86 ChatGPT / 42.53 Gemini) is the study's weakest single persona. US domestic brands have structural AI advantages: local content density, English-language authority sites, and established Wikipedia articles that Veolia's US presence cannot easily replicate in the short term.

T5 — Adjacent technology giants compete for AI attention

Schneider Electric (SOV 5.20% ChatGPT / 7.06% Gemini), Siemens AG (4.55% CG / 8.14% GM), and ABB (6.00% GM) appear regularly in the same AI responses as Veolia for water digitalisation, industrial energy efficiency, and smart infrastructure queries. They are not in Veolia's defined competitor list — but Siemens co-occurs with Veolia in 147 Gemini responses, more than REMONDIS or Waste Management (both defined competitors). These brands compete for AI narrative space in exactly the high-BIS domains — Water Technologies and Bioenergy — where Veolia performs best.

---

SWOT Summary Matrix

	Strengths	Weaknesses
Internal	SOV 21.3% CG / 29.0% GM (2.1–2.6× lead) · BIS +6.99 above pool · Middle East BIS 51.24 · Sentiment at scale (+0.381 CG) · Competitive Score 0.843 GM	Owned media 9.0% CG · US BIS 46.59 (worst market) · Decision gap −1.06 BIS · Bioenergy volume 707 CG · veolia.es 8 CG citations
	Opportunities	Threats
External	Wikipedia: 295 CG citations, one audit rewrites 16.3% of input · veolia.es: 8→280+ citations · Bioenergy scale (quality already at BIS 50.08) · Smart Water Magazine: 106 citations per article · Awareness gap: 2.30 BIS below Consideration	ENGIE BIS 49.18 > Veolia 48.63 CG · SUEZ: 630 GM co-occurrences · Wikipedia uncontrolled (compliance risk) · US domestic density · Siemens/Schneider adjacent competition

Page 20 — GEO Scorecard

---

20.1 All Metrics at a Glance

Metric	Veolia ChatGPT	Veolia Gemini	Pool Avg CG	Pool Avg GM	Rating CG	Rating GM
BIS	48.63	47.31	41.63	37.02	✓ +6.99	✓ +10.29
SOV	21.3%	29.0%	—	—	✓ #1 of 16	✓ #1 of 16
SBOV	22.0%	31.0%	—	—	✓ #1	✓ #1
Position Score	0.269	0.214	0.191	0.107	✓ +0.078	✓ +0.107
Sentiment Score	+0.381	+0.417	0.342	0.361	✓ +0.039	✓ +0.056
Mention Score	0.226	0.154	—	—	✓ #1 CG	✓ #1 GM
Competitive Score	0.777	0.843	—	—	✓ #2 CG	✓ #1 GM
Total Mentions	3,290	4,555	—	—	✓	✓
Responses w/ Veolia	2,066	2,736	—	—	✓	✓
Ranking	#1	#1	—	—	✓	✓

Rating key: ✓ = above market · ✗ = below market · ~ = parity

---

20.2 Metric-by-Metric Analysis

BIS — Brand Impact Score

ChatGPT: 48.63 · Gemini: 47.31

Veolia leads the competitive pool by +6.99 (ChatGPT) and +10.29 (Gemini) BIS points. The only exception: ENGIE records a BIS of 49.18 in ChatGPT — 0.55 points above Veolia. In Gemini, Veolia leads all 16 entities. The BIS gap between engines (1.32 points) is explained primarily by the Mention Score difference: Veolia gets mentioned more deeply per response in ChatGPT than in Gemini.

One-line insight: Veolia is the dominant brand in this competitive set on the composite metric — with one narrow exception (ENGIE, ChatGPT) that signals a specific energy narrative gap.

---

SOV — Share of Voice

ChatGPT: 21.3% (2,066 of 9,374 responses) · Gemini: 29.0% (2,736 of 8,826 responses)

SOV measures how broadly the AI mentions a brand — the percentage of all responses where Veolia appears at least once. Veolia leads by a structural margin: 2.1× Clean Harbors in ChatGPT (9.9%), 2.6× Xylem in Gemini (11.2%). The cross-sector model is the explanation: Veolia qualifies for water, waste, and energy queries simultaneously, while every competitor is relevant in only 1–2 domains.

One-line insight: No other brand comes close to Veolia's AI reach — the lead is structural, not incidental, because no competitor spans all three product lines.

---

SBOV — Share of Branded Voice

ChatGPT: 22.0% · Gemini: 31.0%

SBOV measures Veolia's share within the branded universe — responses where at least one brand is mentioned. At 31.0% in Gemini, Veolia appears in nearly 1 of every 3 competitive comparisons. In ChatGPT (22.0%), the branded landscape is more compressed at the top, but Veolia still holds a commanding first position.

One-line insight: When buyers ask the AI to compare environmental services providers, Veolia is included in 22–31% of all competitive responses — more than any rival.

---

Position Score

ChatGPT: 0.269 · Gemini: 0.214 · Pool avg CG: 0.191 · Pool avg GM: 0.107

Position Score uses logarithmic decay weighting: mention position 0 (first mention) is worth 3.5× more than mention position 10. Veolia's score (+0.078 above pool in ChatGPT, +0.107 in Gemini) reflects that the AI places Veolia early in responses — average mention position 2.26 in ChatGPT, 2.70 in Gemini. Position Score is where Veolia's advantage over the pool is most decisive. Clean Harbors matches Veolia's Position Score in ChatGPT (0.280 vs 0.269) — but across a much narrower market (primarily US hazardous waste).

One-line insight: Veolia is mentioned earlier than competitors — the most cognitively impactful position in any AI response list.

---

Sentiment Score

ChatGPT: +0.381 · Gemini: +0.417 · Pool avg CG: +0.342 · Pool avg GM: +0.361

Sentiment Score (scale −1 to +1) reflects how positively the AI frames Veolia. At +0.381 in ChatGPT, Veolia ranks 6th of 16 on sentiment — positive, but not the highest tone. ENGIE (0.4211), Ameresco (0.4090), and Xylem (0.4001) score higher in ChatGPT sentiment, but with 347, 148, and 1,017 mentions respectively. In Gemini (+0.417), Veolia ranks 2nd — behind Ameresco by 0.0035 points. The hazardous waste product line is the sentiment drag: its compliance-heavy framing (+0.331 CG) pulls below Water Technologies (+0.432 CG) and Bioenergy (+0.428 CG).

One-line insight: Veolia's sentiment is above market at every scale level — the real lever is closing the Hazardous Waste sentiment gap through outcome-focused content.

---

Mention Score

ChatGPT: 0.226 · Gemini: 0.154

Mention Score measures the depth of brand presence per response — how often and how centrally the AI references Veolia within each individual reply. Veolia holds the highest Mention Score among all 16 entities in ChatGPT (0.226). In Gemini, it also leads (0.154), but by a smaller margin. The gap between the two engines (−0.072) is the single largest component explaining the BIS difference. Veolia appears in more Gemini responses (SOV 29% vs 21%), but each Gemini appearance is thinner — fewer mentions per response, less contextual elaboration.

One-line insight: ChatGPT mentions Veolia in fewer responses than Gemini — but each mention is richer and deeper, explaining why ChatGPT BIS is higher despite lower SOV.

---

Competitive Score

ChatGPT: 0.777 · Gemini: 0.843

Competitive Score measures how often Veolia outranks other entities mentioned in the same response. In Gemini, 0.843 is the highest of all 16 entities — Veolia beats 84% of all co-occurring brands per response. In ChatGPT (0.777), Clean Harbors leads with 0.863 — reflecting its more concentrated, single-market (US hazardous waste) profile, where it dominates narrower but more focused query spaces. Veolia's Competitive Score across all 9 product-line × funnel combinations is structurally strong.

One-line insight: In Gemini, Veolia wins the head-to-head comparison in 84% of all competitive co-occurrence scenarios — the highest rate of any brand in the study.

---

20.3 Performance by Product Line

Product Line	CG BIS	CG Mentions	GM BIS	GM Mentions	Assessment
Water Technologies	50.11	986	46.90	1,408	✓ CG leader / ~ GM
Bioenergy & Energy Efficiency	50.08	707	48.39	1,307	✓ CG high / ✓ GM leader
Hazardous Waste	47.15	1,597	46.93	1,840	~ volume · ✗ quality

Inversion pattern: Hazardous Waste generates 49% of all Veolia ChatGPT mentions but produces the lowest BIS. Water Technologies and Bioenergy generate the highest BIS with lower volumes. The AI talks about Veolia more when the topic is waste — but with less enthusiasm.

---

20.4 Performance by Funnel Stage

Funnel Stage	CG BIS	CG Mentions	GM BIS	GM Mentions	Assessment
Consideration	49.78	1,247	48.50	1,693	✓ Best stage both engines
Decision	48.72	1,093	47.36	1,462	~ Solid but below Consideration
Awareness	47.48	950	46.19	1,400	✗ Weakest stage both engines

Pattern: Both engines agree: Consideration > Decision > Awareness. The 1.06-point gap from Consideration to Decision (ChatGPT) is the most actionable single gap in the report.

---

20.5 Performance by Market

Market	CG BIS	GM BIS	Assessment
Middle East	51.24	48.47	✓ Best in both engines
Australia	47.61	47.63	✓ Second, stable
Spain	47.40	46.37	~ Third, owned media gap
United States	46.59	45.53	✗ Worst market, highest competition

---

20.6 Overall GEO Grade

BIS 48.63 (ChatGPT) / 47.31 (Gemini) = Present but not dominant.

The BIS scale interpretation:

BIS Range	Meaning
80+	AI recommends first, spontaneously, confidently
60–80	Well-positioned, consistently visible
40–60	Present but not dominant
<40	Competitors outpace

Veolia sits firmly in the "Present but not dominant" zone — 31 points below where the AI would recommend it first, automatically, and without qualification. The distance is real, but the direction is clear. Veolia is the most visible environmental services brand in generative AI across its defined competitive set. The path from ~48 to ~60 runs through three content investments: Decision-stage depth, Hazardous Waste sentiment elevation, and Gemini mention density.

GEO Overall Grade: A- (Leading the sector, with clear and actionable gaps)

The brand leads on every composite metric except ENGIE's BIS in ChatGPT. It leads on SOV by 2.1–2.6×. It leads on Competitive Score. It leads on Mention Score. No competitor is close on the structural measure that matters most — Share of Voice. The gaps are real but specific, not systemic.

Page 21 — Strategic Imperatives

---

5 Priorities Grounded in Data

Every imperative below identifies a specific measurable problem, a concrete action to address it, and an expected impact. They are ordered by priority — the combination of data evidence, implementation speed, and BIS impact.

---

Imperative 1 — Wikipedia Stewardship

The problem

Wikipedia's "Veolia Environmental Services" article generates 295 ChatGPT citations — the 3rd most-cited individual URL in the entire ChatGPT dataset for Veolia responses. The entire Wikipedia domain accounts for 1,672 ChatGPT citations (16.3% of all source citations in Veolia-mention responses). Veolia's owned media footprint totals 956 ChatGPT citations (9.0%). Wikipedia alone contributes nearly twice as much to Veolia's AI narrative as every Veolia-owned website combined.

Veolia does not control this content. The article is community-maintained. Any historical controversy, regulatory incident, or incomplete information present in that article flows directly into AI responses. The Australia compliance verbatim (page 15) — "Veolia has faced several environmental violations in Australia, including waste and water pollution violations in 2021 and 2024" — is consistent with violation-tracker and Wikipedia-indexed content. The SUEZ merger (completed 2022) is not adequately reflected: SUEZ still appears as an independent competitor in 480 ChatGPT / 630 Gemini co-occurrences.

Action

1. Commission a professional Wikipedia audit of "Veolia," "Veolia Environmental Services," and "SUEZ" articles in English, Spanish, French, and German
2. Verify that the 2022 SUEZ acquisition is accurately reflected — SUEZ positioned as Veolia subsidiary, not independent competitor
3. Update all quantitative data: revenues, employees, countries served, water treated per year, GCC infrastructure projects
4. Add specific project references currently absent: Sadara SWRO, Fujairah 2, Móstoles biomass — each with source citations
5. Prepare primary sources (annual reports, press releases with verifiable data) that Wikipedia editors can reference — do not edit directly

Expected impact

16.3% of ChatGPT's Veolia narrative input improves. No other single action touches this citation volume. Estimated BIS impact: +0.5 to 1.0 points globally in ChatGPT. Secondary effect: SUEZ co-occurrence reduction from 480 to below 350, reducing the "permanent shadow" effect.

---

Imperative 2 — veolia.es Overhaul

The problem

veolia.es generates 8 ChatGPT citations across 7 unique URLs — despite Spain being the 3rd-largest market in this study and a competitive battleground for both hazardous waste (Tradebe: 198 co-occurrences, TMA: 61 CG / 332 GM citations) and water (Aqualia, Acciona). Comparable regional domains: anz.veolia.com generates 144 ChatGPT / 724 Gemini citations across 55 / 123 URLs. near-middle-east.veolia.com generates 108 ChatGPT / 314 Gemini across 22 / 38 URLs — at the highest citation-per-URL efficiency of any Veolia domain (8.3 in Gemini). veolia.es achieves 1.1 citations per URL in ChatGPT.

The Spain BIS gap is a direct consequence: 47.40 ChatGPT / 46.37 Gemini — 3rd of 4 markets, with Gemini showing the largest inter-engine deterioration (Spain trails Australia by 1.26 BIS points in Gemini vs. 0.21 in ChatGPT). The Spanish AI narrative is written by Tradebe, TMA, and Spanish market research firms, not by Veolia.

Action

1. Rebuild veolia.es content depth to match veolianorthamerica.com standards: - Target: 80+ indexed URLs (currently ~20 estimated) - Content format: outcome-focused case studies with measurable results, service specification pages, Spanish regulatory certifications, NRW reduction case studies with investment data (the €360M Spain investment figure from page 15 is already AI-indexed — expand it)
2. Add FAQ sections targeting Decision-stage queries in Spanish: certifications, compliance records, capacity data for hazardous waste treatment
3. Structured data: Schema.org/Organization with disambiguatingDescription, Schema.org/Service for each service line, local business data for Spain operations
4. Publish Spanish-language differentiation content vs. Tradebe: scale, technology portfolio, circular economy outcomes

Expected impact

If veolia.es reaches 3.5 citations per URL across 80 indexed URLs: ~280 ChatGPT citations — from 8 today. Spain BIS improvement: estimated +1.5 to 2.5 points. Moving Spain from 3rd to 2nd market by BIS would close the gap with Australia (currently 47.61 CG).

---

Imperative 3 — Decision-Stage Content Depth

The problem

Decision stage produces the second-lowest BIS of all three funnel stages: 48.72 ChatGPT / 47.36 Gemini. The gap from Consideration (the best stage) to Decision: −1.06 ChatGPT / −1.14 Gemini. Both engines agree on the ranking: Consideration > Decision > Awareness. Decision-stage sentiment (+0.364 ChatGPT) is the lowest of all three stages — 0.028 points below Consideration.

The mechanism is clear: Decision-stage queries ask for specific validation — "What certifications does Veolia hold?", "Show me a US case study for hazardous waste treatment", "What outcomes has Veolia achieved in water reuse?" The AI answers with whatever specific, verifiable content it has indexed. Veolia's Decision-stage content is thinner than its Consideration-stage brand recognition. The AI knows Veolia is a strong option (Consideration) but cannot fully validate it when buyers are ready to act (Decision).

The named product gap amplifies this: when Gemini compares digital water platforms, SUEZ's "AQUADVANCED®" gives it a specificity advantage that Veolia's unnamed equivalent cannot match. Named technology documentation beats unnamed capabilities in Decision-stage AI responses.

Action

1. Publish a library of 10–15 outcome-focused case studies in AI-indexable format: specific client, geography, measurable result (volume processed, efficiency %, cost saved, emissions reduced) — following the model of the "1.5 billion gallons / 84% efficiency" verbatim (page 15) that generates 5/5 sentiment and position 0
2. Create certification and accreditation pages per market: US (EPA compliance status, ISO certifications), Australia (EPA Queensland/Victoria status, NATA accreditation), Spain (Registro de Actividades de Gestión de Residuos), Middle East (UAE TRA, Saudi Aramco approvals)
3. Name Veolia's digital water platform explicitly in content: following the SUEZ/AQUADVANCED® model, a named product with documented specifications competes better at Decision stage
4. Post SUEZ integration content: explicitly document that AQUADVANCED® capabilities are now Veolia Water Technologies assets

Expected impact

Close the 1.06-point BIS gap between Consideration and Decision in ChatGPT. Target: Decision-stage BIS above 50.0 in both engines within 12 months. Secondary effect: Australia compliance narrative managed — a Veolia-owned "Environmental Performance" page competes with violation-tracker citations at Decision stage.

---

Imperative 4 — Bioenergy Content Activation

The problem

Bioenergy & Energy Efficiency has the highest ChatGPT position score (0.293) and the highest Gemini BIS (48.39). ChatGPT BIS: 50.08 — essentially tied with Water Technologies (50.11) for the top product line. Yet ChatGPT mentions: 707 — the fewest of all three product lines. Hazardous Waste: 1,597 mentions. Water Technologies: 986.

This is a pure volume gap. The AI already knows Veolia is the leading Bioenergy operator — it picks Veolia first for district heating queries (Móstoles verbatim), biomethane queries ("one of the most established operators globally"), and waste-to-energy assessments. The quality ceiling is already reached. But the market of potential buyers asking the AI about bioenergy is capturing Veolia at 707 mentions while the waste-focused market gets 1,597.

The ENGIE proximity in ChatGPT (BIS 49.18 vs Veolia 48.63) is strongest in energy and bioenergy topics — the verbatim in page 7 shows ChatGPT treating Veolia and ENGIE as "co-leaders" in biomethane. More Veolia-specific Bioenergy content would differentiate this positioning.

Action

1. Publish a minimum of 8 Bioenergy case studies with measurable outcomes: biogas yields (m³/day), biomethane injection volumes, district heating capacity (MW), CO₂ avoided per project
2. Create a "Bioenergy Projects Portfolio" page on veolia.com and near-middle-east.veolia.com with all current active projects — matching the Middle East water infrastructure project references that drive BIS 51.24 in that market
3. Submit 2–3 articles to Smart Water Magazine specifically about Bioenergy topics (current Smart Water Magazine coverage is water-focused; the 430 GM citation volume would amplify Bioenergy content with exceptional ROI)
4. Produce a biomethane technology explainer page on veoliawatertech.com with differentiation vs. ENGIE: scale, integration with waste management, circular economy closed loop

Expected impact

Scale Bioenergy ChatGPT mentions from 707 to 1,000+, approaching Water Technologies volume. Estimated ChatGPT BIS for Bioenergy: 50.5+ (currently 50.08). Secondary effect: reduce ENGIE's narrative overlap in energy topics, contributing to ENGIE BIS containment (Imperative 5).

---

Imperative 5 — ENGIE Containment in ChatGPT

The problem

ENGIE BIS 49.18 vs. Veolia BIS 48.63 in ChatGPT — the only instance in the entire study where a competitor leads Veolia on the composite metric. The gap: 0.55 BIS points, driven by ENGIE's advantage across three sub-components simultaneously: Sentiment (+0.4211 vs 0.3806 = +0.04 advantage), Position Score (+0.2803 vs 0.2689 = +0.01), Mention Score (+0.2395 vs 0.2256 = +0.01). Veolia beats ENGIE only on Competitive Score (0.7768 vs 0.7414).

The mechanism: in energy efficiency and bioenergy queries, ENGIE's concentrated energy identity generates more pointed, enthusiastic AI responses than Veolia's broader multi-sector positioning. The AI treats Veolia and ENGIE as co-leaders in bioenergy (page 15 verbatim); in that context, ENGIE's deeper energy narrative narrowly wins the BIS comparison. In Gemini, this reverses: Veolia (47.31) leads ENGIE Middle East (44.55) by 2.76 points — the ChatGPT alert is real but narrow. ENGIE's SOV in ChatGPT is 2.40% vs. Veolia's 21.26% — it leads on quality in one narrow domain, not on breadth.

Action

1. Produce content explicitly comparing Veolia and ENGIE in Bioenergy and district heating — emphasising Veolia's integrated model (waste-to-energy + water management + industrial efficiency in one contract) vs. ENGIE's standalone energy focus
2. Publish a "Veolia vs. alternatives" positioning page for industrial decarbonisation queries: Veolia offers energy efficiency as part of an integrated environmental contract; ENGIE offers energy services standalone
3. Target Smart Water Magazine and WasteToEnergyEurope.org for Veolia Bioenergy coverage — the specific publications where ENGIE's energy narrative currently dominates Veolia's in ChatGPT training sources
4. Name Veolia's energy efficiency service brands: follow ENGIE's model where product-level specificity (named services, named technology platforms) generates higher AI mention depth and sentiment

Expected impact

Close the 0.55-point ENGIE–Veolia BIS gap in ChatGPT. Target: Veolia BIS above 49.0 in ChatGPT within 12 months, creating a clear margin above ENGIE across both components. The Bioenergy content produced for Imperative 4 directly feeds Imperative 5 — the two share content assets.

---

Priority Summary

#	Imperative	Problem (data)	Timeline	BIS Impact
1	Wikipedia stewardship	16.3% CG narrative uncontrolled · 295 CG citations	0–3 months	+0.5–1.0 global
2	veolia.es overhaul	8 CG citations vs 144 for anz.veolia.com	1–6 months	+1.5–2.5 Spain
3	Decision-stage content	−1.06 BIS at Decision vs Consideration (CG)	2–6 months	+1.0 Decision stage
4	Bioenergy activation	BIS 50.08 but 707 CG mentions (vs 1,597 Hazwaste)	2–5 months	+0.5 Bioenergy line
5	ENGIE containment	ENGIE BIS 49.18 > Veolia 48.63 in CG	3–6 months	+0.5 CG overall

Page 22 — 90-Day Roadmap

---

Structure: 3 Sprints, Ordered by Impact and Feasibility

The roadmap translates the 5 strategic imperatives into a sequenced 90-day execution plan. Sprint 1 prioritises low-cost, high-impact actions that require audit and editorial work — no new infrastructure. Sprint 2 requires content creation. Sprint 3 requires distribution, publication, and re-measurement.

---

Sprint 1 — Foundation and Audit (Days 1–30)

Objective: Audit the full GEO footprint, fix structural vulnerabilities, and prepare the content brief.

Action	Owner	Data Justification	Deliverable
Wikipedia audit (EN/ES/FR/DE) — Veolia + SUEZ articles	External Wikipedia consultant	295 CG citations for Veolia article · 16.3% CG narrative · SUEZ still treated as independent competitor (480 CG co-occurrences)	Audit report + list of corrections with primary sources
veolia.es content audit — URL inventory, gap map	SEO/GEO team	8 CG citations vs 144 (anz) · 2.2 citations/URL vs 4.7 for veolianorthamerica.com	Content brief: 80-URL target, priority pages, format templates
S.A.M. alignment scan on veolia.es, veolianorthamerica.com, veoliawatertechnologies.com	GEO analyst	Identify which Decision-stage prompts no existing Veolia page addresses	Gap report: top 25 unanswered Decision queries per market
Schema markup audit — Organization, Service, FAQPage	Dev/SEO team	Wikipedia wiki structure gives 16.3% narrative; Schema gives AI structured data from owned pages	Schema implementation plan for all 6 primary domains
anz.veolia.com compliance page brief	Content team (ANZ)	Australia compliance verbatim (page 15): violations 2021/2024 cited at Decision stage by ChatGPT	Page brief: "Environmental Performance and Standards" for services.veolia.com.au
Smart Water Magazine outreach — partnership proposal	PR/comms	430 GM citations from 349 responses · Single article: 106 citations	Content partnership agreement and editorial calendar
InsightDesk/GeoRadar baseline configuration	Analytics team	Establish T0 metrics across all KPIs before any content goes live	Live dashboard with BIS, SOV, funnel, market, product line by engine

Sprint 1 output (Week 4): - Wikipedia audit report with specific corrections and primary source references ready to submit - veolia.es 80-URL content brief approved - S.A.M. gap map: top Decision-stage query gaps per market - Schema.org implementation deployed on veolia.es, veolia.com, veolianorthamerica.com - Smart Water Magazine editorial calendar confirmed (minimum 4 articles/year) - T0 dashboard live

---

Sprint 2 — Content Creation (Days 31–60)

Objective: Produce the core content assets that close the three primary BIS gaps: Decision stage, Spain, and Bioenergy.

Action	Owner	Data Justification	KPI Target
10 Decision-stage case studies (EN)	Content team	Decision BIS 48.72 CG — needs specific, outcome-measurable content · "1.5bn gallons / 84% efficiency" verbatim proves format works	10 published pages across veolianorthamerica.com, anz.veolia.com, near-middle-east.veolia.com
veolia.es — 20 priority pages (Phase 1)	Spain content team	8 CG citations today · target 80+ URLs at 3.5 citations/URL = 280 CG citations	20 pages live: 8 case studies (ES), 8 service pages, 4 FAQ/compliance
Certification and accreditation pages (US, AU, ES, ME)	Legal + content team	Decision gap − compliance queries return violation-tracker citations; owned certification pages would compete	4 pages live (one per market) with full accreditation inventory
Bioenergy case studies — 4 projects with outcome data	Bioenergy content team	BIS 50.08 with 707 CG mentions (vs 1,597 Hazwaste) — volume gap, quality already proven	4 published: 1 per market, measurable outcomes (CO₂ avoided, MW thermal, m³/day biogas)
SUEZ integration content series — 3 articles	comms + tech team	suez.com still cited 406 GM times · AQUADVANCED® name-specific advantage in Decision responses	3 articles on veolia.com + veoliawatertechnologies.com: "SUEZ capabilities within Veolia"
Smart Water Magazine — Article 1 (Water Tech/GCC)	PR team	106 GM citations for single GCC article already · Bioenergy and Decision-stage topics have zero equivalent	Article submitted: GCC water + Bioenergy decarbonisation, Veolia-attributed outcome data
Digital water platform naming content	Product marketing	SUEZ AQUADVANCED® specificity advantage at Decision stage · Unnamed Veolia equivalent loses comparison	Named platform page on veoliawatertechnologies.com with specs, integrations, client references
Hazardous Waste sentiment content (outcomes frame)	Content team	HazWaste sentiment +0.331 CG vs Water Tech +0.432 — compliance framing drags the score	5 HazWaste case studies reframed from compliance to circular economy outcomes

Sprint 2 output (Week 8): - 20 veolia.es pages live (Phase 1) - 10 Decision-stage case studies published across regional domains - 4 Bioenergy case studies published - Certification pages live for all 4 markets - First Smart Water Magazine article submitted - SUEZ integration series (1 of 3) published - Named digital water platform page live

---

Sprint 3 — Scale and Measure (Days 61–90)

Objective: Amplify Bioenergy and US content, distribute Sprint 2 assets for maximum indexability, and measure the first T1 delta.

Action	Owner	Data Justification	KPI Target
veolia.es — 20 priority pages (Phase 2)	Spain content team	Phase 1 baseline establishes citation pattern; Phase 2 scales to 40+ URLs	40+ pages total live on veolia.es
Bioenergy content push — US competitive response	Content + PR team	US BIS 46.59 CG (worst market) · Urban Energy Planning Advisor US is lowest persona (BIS 42.53 GM) · ENGIE targets same US energy queries	4 US-specific Bioenergy/efficiency articles; veolianorthamerica.com energy section rebuild
Smart Water Magazine — Article 2 (Bioenergy/Spain)	PR team	Bioenergy 707 CG mentions vs 1,597 HazWaste · Spain Bioenergy projects (Móstoles) already AI-referenced but underscaled	Submitted: Bioenergy integration in European cities — Spain and France case data
Wikipedia corrections submitted	Wikipedia consultant	Corrections prepared in Sprint 1, community review cycle takes 4–8 weeks	Primary sources submitted · Corrections live or in review
SUEZ integration series — Articles 2 and 3	Comms team	SUEZ still at 406 GM co-occurrences · Post-merger framing absent from AI responses	Both articles published: water tech and waste treatment integration
Press distribution — businesswire.com / Globe Newswire	PR team	businesswire.com and globenewswire.com already indexed by ChatGPT in Veolia responses	3–5 press releases with outcome data from Sprint 2 content, citing veolia.es
GeoRadar re-run (same prompts, same personas)	Analytics team	Establish T0→T1 comparison under controlled conditions	Full re-run: 9,720 prompts, CG + GM, all 36 personas
T0→T1 analysis and roadmap recalibration	GEO analyst	Measure which actions moved which metrics	Delta report: BIS by market / product line / funnel / persona vs T0

Sprint 3 output (Week 12): - veolia.es: 40+ pages live - GeoRadar re-run completed - T0→T1 delta report: BIS, SOV, owned media %, funnel by engine - Sprint 4 priorities defined (12-month roadmap)

---

KPI Target Table

KPI	T0 (Current)	Sprint 1 Target (Day 30)	Sprint 2 Target (Day 60)	Sprint 3 Target (Day 90)
BIS ChatGPT (global)	48.63	48.63 (baseline locked)	—	49.5+
BIS Gemini (global)	47.31	47.31 (baseline locked)	—	48.0+
BIS Spain (ChatGPT)	47.40	—	48.0	49.0+
BIS US (ChatGPT)	46.59	—	—	47.5+
BIS Decision stage (ChatGPT)	48.72	—	—	49.5+
BIS Bioenergy (ChatGPT)	50.08	—	—	50.5+
SOV ChatGPT	21.3%	Maintain	Maintain	22.0%+
SOV Gemini	29.0%	Maintain	Maintain	29.5%+
Owned media % (ChatGPT)	9.0%	—	10.5%	12.0%+
Owned media % (Gemini)	7.4%	—	8.5%	10.0%+
veolia.es citations (ChatGPT)	8	—	50+	120+
Smart Water Magazine (GM citations)	430	430	450+	500+
SUEZ co-occurrences (CG)	480	—	—	<400

---

GEO Tool Suite

Tool	Sprint	Role
S.A.M. (Semantic Alignment Machine)	Sprint 1	Map prompt coverage gaps on existing Veolia pages — identify which Decision queries go unanswered
GEOdoctor	Sprint 1–2	Schema markup implementation, technical GEO audit, FAQ structure optimisation
InsightDesk	Sprint 1–3	Real-time monitoring dashboard — track citation volume and source type changes as content goes live
GeoRadar	Sprint 3	Full re-run to measure T0→T1 BIS delta under controlled conditions
GEODesk AI	Sprint 3	Semantic analysis of delta — which content types moved which metrics, prioritisation for Sprint 4

Page 23 — Conclusion

---

What We Know

9,720 neutral prompts. 18,200 AI responses. 7,845 spontaneous Veolia mentions across ChatGPT and Gemini.

No prompt asked the AI to talk about Veolia. It did so on its own initiative — because when a procurement manager in Dubai asks ChatGPT about industrial water treatment, when a sustainability director in Sydney asks Gemini about bioenergy, or when a waste authority in Madrid queries an LLM about hazardous waste operators, the AI decides to include Veolia in the answer. 3,290 times in ChatGPT. 4,555 times in Gemini.

The numbers confirm it: SOV 21.3% in ChatGPT, 29.0% in Gemini. Veolia appears in 1 of every 5 ChatGPT responses in its market space, and in nearly 1 of every 3 in Gemini. No other environmental services brand in this study comes close — the nearest competitor is 2.1× to 2.6× further down the ranking. Veolia's BIS of 48.63 (ChatGPT) and 47.31 (Gemini) outpaces the competitor pool by +6.99 and +10.29 points respectively.

Sentiment is positive in both engines (+0.381 CG / +0.417 GM). Zero mentions below sentiment level 1 in ChatGPT. The Middle East market delivers BIS 51.24 — the AI talks about Veolia's Gulf infrastructure projects with authority and specificity. Water Technologies (BIS 50.11 CG) and Bioenergy (BIS 50.08 CG / 48.39 GM) represent the brand at its best: the AI treats Veolia as a reference operator, not a list entry.

---

What the Data Reveals

1. Veolia leads — but has not yet crossed into dominance

A BIS of ~48 means the AI mentions Veolia consistently, often early, and with positive framing. It does not yet mean the AI recommends Veolia first, automatically, and without qualification. That threshold is BIS 70+. The gap is 31 points — not insurmountable, but real. Brands cross it by moving from broad recognition (Consideration stage, where Veolia scores 49.78 CG) to specific, trusted recommendation (Decision stage, where Veolia scores 48.72 — 1.06 points lower). That single gap is the most concrete bottleneck in the entire report.

When a buyer is ready to act and asks the AI to validate Veolia specifically — certifications, project references, outcome data, compliance track record — the AI does not have enough specific, citable material to answer with full confidence. The result: Veolia drops out of pole position and the response hedges. The fix is not brand-building. It is content: 10–15 outcome-focused case studies with measurable results, published in AI-indexable format across regional domains.

2. Veolia's AI narrative is not Veolia's to control — yet

9.0% of ChatGPT citations in Veolia-mention responses come from Veolia-owned domains. Wikipedia controls 16.3% — nearly twice as much. The 295 ChatGPT citations for the "Veolia Environmental Services" Wikipedia article make it the single most powerful document shaping what the AI knows about Veolia. Competitor sites contribute ~4.9% — meaning rivals are authoring nearly half as much of the Veolia narrative as Veolia itself. The Spain domain (veolia.es) generates 8 ChatGPT citations. A comparable regional domain (anz.veolia.com) generates 144. The content gap in Spain is 18× in size.

None of this is permanent. Every citation gap is a content investment opportunity. Wikipedia can be stewarded through primary source contribution. veolia.es can be rebuilt. Market research firms can be seeded with outcome data. Smart Water Magazine — which already generates 430 Gemini citations from a single editorial relationship — can be engaged as a recurring partner.

3. The structural advantage is real and defensible — with one exception

ENGIE is the only competitor that leads Veolia on BIS, and only in ChatGPT: 49.18 vs 48.63. The margin (0.55 points) is narrow, but the mechanism is instructive: ENGIE's concentrated energy identity produces more enthusiastic AI responses in bioenergy and energy efficiency contexts. Veolia and ENGIE are treated as co-leaders in biomethane queries. In Gemini, this reverses — Veolia (47.31) leads ENGIE Middle East (44.55) by 2.76 points. The ChatGPT energy narrative is where Veolia needs sharpening. Bioenergy content volume (currently 707 CG mentions) is the lever — the quality is already there (BIS 50.08), only the volume needs scaling.

4. The efficiency insight: not all volume is equal

Hazardous Waste generates 49% of all Veolia ChatGPT mentions (1,597 of 3,290) — yet produces the lowest BIS (47.15) and the lowest sentiment (+0.331) of the three product lines. Bioenergy generates the fewest mentions (707) but the highest position score and near-equal BIS to Water Technologies. The investment implication: resources should not flow to the highest-volume product line. They should flow to the highest-efficiency lines (Bioenergy, Water Technologies) and to addressing the Hazardous Waste sentiment gap through narrative reframing — from compliance language to circular economy outcomes.

---

What Comes Next

Three actions with the highest probability of moving the BIS before the next audit:

1. Wikipedia. Commission the audit immediately. 295 citations in one article. 16.3% of the ChatGPT narrative footprint. One structured intervention improves the quality of the AI's primary reference document for Veolia — reaching every subsequent response that draws on it.

2. veolia.es. Rebuild it with the depth of veolianorthamerica.com. 8 citations today could become 280+ in 6 months by applying the same content density that already works for Veolia's North American domain. Spain is a competitive market (Tradebe, TMA, Aqualia all appear in the AI's Spanish responses alongside Veolia). The brand presence there should match the commercial stakes.

3. Decision-stage content. Publish case studies that the AI can cite as specific validation. The verbatim from page 15 shows what works: "1.5 billion gallons saved, 84% network efficiency, $1.1 million in savings" — a response the AI generated because veolianorthamerica.com published specific, measurable outcome data. Every regional market needs 3–5 equivalent case studies for the AI to draw from at the moment when buyer intent is highest.

---

The Number That Matters

Veolia holds BIS 48.63 in ChatGPT and 47.31 in Gemini — leading all 15 defined competitors by +6.99 and +10.29 points respectively. It is #1 by Share of Voice in both engines with a 2.1–2.6× margin.

The gap to dominance is 31 BIS points. The path runs through three content investments — not brand investment — that are executable in 90 days.

---

March 2026

498 Advance — GEO Intelligence & Brand Visibility

Report generated with GeoRadar — Generative AI Visibility Audit Engine