Exclusive Next40 study analyzing AI search performance across 4 major platforms. Discover which French startups dominate generative engine visibility and the strategic framework to win the AI search race.

Publication Date: July 18, 2025 Last Updated: July 18, 2025

TL;DR - Executive Summary

The reality: 1.6 billion users now make purchase decisions through AI conversations. Our exclusive Next40 analysis reveals a critical gap between AI brand visibility and business performance.

Key findings:

53% average visibility across AI platforms

69% for B2C vs 44% for B2B companies

61% of companies in suboptimal categories

90% reduction in traditional click-through rates

Immediate action: Test your AI visibility at getmint.ai and implement the 3 priority levers:

external source optimization,

structured content,

and regular updates.

📋 Table of Contents

The end of Search monopolistic era

Mint Methodology: how we mapped the AI Search landscape

53% average visibility: many Next40 Companies are not getting their faire share in AI Search

Complete Performance Rankings

Four Categories of AI Search Performance

How AI Models Actually Choose Companies

What about content optimization for LLM? Going beyond SEO, aim for LLM extractability

The Mint Strategic Framework for AI Search Dominance

The business impact: why this matters now

The Inflection Point: Act Now or Fall Behind

Next40 Startup Study

AI is no longer just a trend, it’s a tectonic shift in how customers search, evaluate, and choose solutions.

As nearly half of consumers now rely on AI models like ChatGPT, Gemini, Perplexity, or Mistral to guide their purchase decisions, businesses face a new form of discoverability: visibility within generative AI responses.

In our exclusive Les Echos x Mint study, we analyzed the presence and positioning of the Next40 startups across 4 major AI platforms. The findings reveal a sharp contrast between AI brand perception and AI visibility, and underscore the urgent need to go beyond traditional SEO, content strategy, and influence ecosystems.

This report is not a ranking, but a diagnostic tool. It highlights the structural shifts already underway and identifies where early movers are gaining ground, not just in clicks, but in conversational authority, qualified trafic, and long-term competitive advantage.

The end of Search monopolistic era

The Fundamental Shift

While companies are still focused on optimizing solely for Google rankings, customers are quietly but firmly migrating to AI chatbots. 1.6 billion monthly users across ChatGPT, Gemini, Perplexity and Claude now make purchase decisions through conversation, not clicks. 5 billion ChatGPT visits in April 2025 alone demonstrate the scale of this transformation.

The death of the click model: Traditional search generated 25-30% click-through rates (CTR). AI search delivers 3-4% CTR, a 90% reduction. Your customers are getting answers without (almost) ever visiting your website.

See where and how your company shows when customers ask AI for recommendations
Get My Free AI Visibility Score →

Why this marks a radical change in Search behaviors and capabilities

Query evolution reveals intent shift:

Traditional search: 4 words average ("accounting software pricing")

AI search: 20-25 words average ("I need an accounting software for a 50-person SaaS company with international subsidiaries, what do you recommend?")

Session depth indicates decision quality:

Traditional search: 30 seconds per query

AI search: 6 minutes per conversation

This isn't just a new channel, it's a fundamental shift in how buyers discover, evaluate, and choose solutions.

Mint Methodology: how we mapped the AI Search landscape

Study Architecture

Our partnership with Les Echos delivers the first comprehensive AI search analysis of France’s top tech startups:

Scope and Scale:

40 tech / deeptech scale-ups analyzed across all major AI platforms (Next40 laureates)

4,500 AI responses evaluated using industry-standard prompts

25,000+ unique sources consulted by AI models

3,000 unique domains in the AI source ecosystem

AI Models Tested:

ChatGPT 4o (OpenAI)

Gemini Flash 2.0 (Google)

Mistral Medium 2025

Perplexity Sonar Pro

The Mint Scoring Framework : New Key Success Metrics for AI-Based Visibility

We developed a proprietary 3-dimension scoring system that correlates with actual brand existence throughout models:

VISIBILITY: Brand mention rate across AI models’ responses

PERCEPTION: Sentiment analysis equivalent to generative NPS

CONTENT: Content score along 3 main dimensions (Technical, Structure, Quality) and 11 sub-parameters

53% average visibility: many Next40 Companies are not getting their faire share in AI Search

B2C companies achieve 69% average visibility, appearing in 7 out of 10 relevant AI responses.

B2B companies manage only 44% visibility, appearing in fewer than half of relevant searches.

The platform disparity: Gemini delivers 67% visibility rates while Mistral shows 47%.

Complete Performance Rankings

Les Echos / Mint (getmint.ai) Next40 AI Search Index - 15 B2C companies

Rank	Company	Market	Visibility Score	Perception Score	Content Score	Overall Score
1	Qonto	🇫🇷 FR	95%	58%	63%	72%
2	Ekwateur	🇫🇷 FR	88%	58%	62%	69%
3	Back Market	🇫🇷 FR	93%	50%	59%	67%
4	Doctolib	🇫🇷 FR	99%	58%	43%	67%
5	Le Hibou	🇫🇷 FR	100%	42%	54%	65%
6	Malt	🇫🇷 FR	100%	58%	38%	65%
7	Spendesk	🇫🇷 FR	68%	68%	47%	61%
8	Mistral AI	🇬🇧 Global	72%	58%	45%	58%
9	BlaBlaCar	🇫🇷 FR	100%	8%	29%	46%
10	Ledger	🇬🇧 Global	36%	42%	53%	44%
11	Mistertemp'	🇫🇷 FR	18%	67%	42%	42%
12	Agryco	🇫🇷 FR	17%	50%	51%	39%
13	Gojob	🇫🇷 FR	16%	42%	50%	36%
14	ClubFunding	🇫🇷 FR	32%	25%	28%	28%
15	Vestiaire Collective	🇬🇧 Global	98%	25%	N/A	N/A

Les Echos / Mint (getmint.ai) Next40 AI Search Index - 25 B2B companies

Rank	Company	Market	Visibility Score	Perception Score	Content Score	Overall Score
1	Brevo	🇬🇧 Global	88%	92%	60%	80%
2	Innovafeed	🇬🇧 Global	96%	67%	66%	76%
3	Weezevent	🇫🇷 FR	79%	83%	61%	74%
4	PayFit	🇫🇷 FR	88%	67%	57%	71%
5	Pennylane	🇫🇷 FR	86%	67%	57%	70%
6	Swile	🇫🇷 FR	93%	75%	39%	69%
7	Exotec	🇬🇧 Global	45%	92%	63%	67%
8	Flying Whales	🇬🇧 Global	87%	50%	59%	65%
9	EcoVadis	🇬🇧 Global	88%	42%	43%	58%
10	Contentsquare	🇬🇧 Global	56%	67%	47%	57%
11	Descartes Underwriting	🇬🇧 Global	45%	75%	50%	57%
12	Alan	🇫🇷 FR	32%	75%	56%	54%
13	Electra	🇫🇷 FR	29%	67%	56%	51%
14	ChapsVision	🇫🇷 FR	12%	67%	65%	48%
15	Alice & Bob	🇬🇧 Global	11%	67%	63%	47%
16	Verkor	🇬🇧 Global	12%	67%	55%	45%
17	NW Groupe	🇫🇷 FR	7%	67%	55%	43%
18	Homa	🇬🇧 Global	26%	58%	42%	42%
19	Pigment	🇬🇧 Global	1%	75%	48%	41%
20	Qair	🇬🇧 Global	0%	67%	57%	41%
21	Mirakl	🇬🇧 Global	20%	58%	46%	41%
22	Zeplug	🇫🇷 FR	20%	33%	68%	40%
23	Voodoo	🇬🇧 Global	69%	17%	34%	40%
24	Equativ	🇬🇧 Global	12%	58%	48%	39%
25	Worldia	🇫🇷 FR	2%	58%	41%	34%

Discover your company's AI search performance across all major platforms Get Your Complete AI Visibility Audit →

Four Categories of AI Search Performance

Visibility × Perception positioning of B2C members of the Next40

Visibility × Perception positioning of B2B members of the Next40

Strategic Positioning Analysis

Cluster	Definition	B2B (10/25 companies)	B2C (7/15 companies)	Editorial Angle / Key Insight
Superstars	Visibility ≥ average Perception ≥ average	- Swile - Brevo - Innovafeed - PayFit - Pennylane - Weezevent	- Malt - Doctolib - Ekwateur - BackMarket - Qonto	Next40 members proving that strong visibility can align with strong brand positioning
Polarizing Giants	Visibility ≥ average Perception < average	- EcoVadis - Flying Whales	- BlaBlaCar - Le Hibou - Vestiaire Collective	Huge visibility but fragile perception
Hidden Gems	Visibility < average Perception ≥ average	- Pigment - Qair - NW Groupe - Alice & Bob - ChapsVision	- Mistertemp’ - Agryco	Excellent perception in AI tools but still low end-user visibility → strong opportunity space
Underperformers	Visibility < average Perception < average	- Zeplug - Voodoo	- ClubFunding - Gojob - Ledger	Need to prioritize fixing brand perception

Note: The 2 remaining B2C companies and 8 B2B companies fall into a neutral zone (visibility and/or perception close to the average), which we have excluded here to maintain clarity in the clustering.

The Hidden Gems Opportunity

61% of Next40 companies fall into suboptimal categories, creating massive competitive repositioning opportunities. Hidden Gems represent the highest ROI potential: companies like Pigment (1% visibility, 75% perception) and ChapsVision (12% visibility, 67% perception) have strong AI sentiment foundations requiring only visibility amplification.

How AI Models Actually Choose Companies

The Source Ecosystem Reality

Critical finding: Your website represents only 10-20% of your AI search presence. The remaining 80-90% depends on external ecosystem mentions across:

Other companies: 68% of visibility sources (searching for competitors, alternatives, etc.)

Industry directories: 11% of visibility sources

Community platforms: 35% of perception sources

Review sites: 24% of perception sources

Platform-Specific Source Preferences

Highly scattered sources allocation overall (top 10 sources capturing only c.15% of total sources consultation) indicating a need for deeply granular analysis and wide influence potential over a long tail of external websites

ChatGPT's authority bias: 14.1% of citations from Wikipedia, 3.4% from Reuters, indicating preference for established, authoritative sources.

Gemini's directory focus: 1.5% from Appvizer.fr, 0.8% from Codeur.com, showing preference for specialized business directories.

Perplexity's balanced approach: More evenly distributed sources with 1.3% from Appvizer.fr, 1.0% each from YouTube and Shopify.

Reddit: A global influence, but missing from the French AI landscape

Despite ranking #19 in English AI perception sources, Reddit is only ranking #63 for the French perception context.

This contrast highlights a critical geographic asymmetry: while Reddit plays a key role in shaping brand narratives in global LLMs and in English, it remains under-leveraged or almost ignored so far by French digital ecosystems, limiting its impact on AI-generated results for French users.

For companies aiming for international / USA AI visibility, Reddit truly deserves strategic attention.

Given the very recent growth focus put by Reddit on the French market, it appears wise to take a longer-term view and to start contributing to the narrative in local languages.

The Competitive Density Problem

The 8-company average per response creates what we call "competitive density", even appearing in AI results doesn't guarantee prominence. Gemini mentions 10 companies per response versus OpenAI's 6, fundamentally changing competitive dynamics.

Strategic implication: Differentiation within AI responses matters more than just getting mentioned. Companies need distinctive positioning that makes them memorable in crowded AI-generated lists.

Get a detailed analysis of what's blocking your AI search visibility Analyze My Content for AI Search →

What about content optimization for LLM? Going beyond SEO, aim for LLM extractability

The Three-Layer Performance Drop

The AI content readiness gap: Analysis of 3,900 company pages reveals a systematic performance decline:

Technical Criteria (B2B: 68%, B2C: 62%): Most companies maintain solid technical foundations, structured data, clean HTML, optimized URLs. This represents baseline competency, not competitive advantage.

Content Structure (B2B: 62%, B2C: 52%): The gap widens as AI-specific requirements diverge from traditional SEO. Companies excel at meta descriptions but fail at content freshness indicators and question-based headings.

Content Depth (B2B: 30%, B2C: 29%): The critical failure point where competitive advantage emerges. This 33-point drop from technical scores to content depth reveals where most companies fail, and where leaders separate themselves.

The 11-Parameter Content Analysis

Our proprietary content analysis weighted AI-critical factors:

3x multiplier factors:

Content freshness indicators

2x multiplier factors:

Main Heading (H1) structure and content

Question-based subheadings and structure

Standard factors:

Comparison content formats

In-depth guide structure

Citation authority

Clean HTML structure (Javascript not performing well with LLM)

Structured data implementation

Meta descriptions

Image alt attributes

URL structure

Get your personalized AI search optimization roadmap Start My Free AI Visibility Analysis →

The Mint Strategic Framework for AI Search Dominance

Concept Ownership + Sources Ecosystem Optimization + LLM-friendly Content

Be the Concept Owner

Core principle: AI models prefer citing definitive experts through comprehensive and authoritative content over multiple average and more similar content inputs sources.

Implementation requirements:

Unique perspective development

Proprietary data integration

Cross-format content deployment

High-frequency publishing with editorial quality

Source Ecosystem Optimization

External optimization priority matrix for Next40 companies

Tier 1 (Immediate impact):

French market: Appvizer.fr, Tool-advisor.fr (directories have a direct impact on Visibility)

Global market: Wikipedia, Gartner

Perception management: G2, Glassdoor, Tripadvisor

Many long tail and smaller websites where short-term influence value truly lies today → to be adapted to your market/business priorities!

Tier 2 (Strategic development):

Community platforms: Reddit

Industry publications: Sector-specific media

Social platforms: LinkedIn, YouTube

Authoritative PR (especially in English today)

AI-First Content optimization priorities

Priority 1 (high impact, low effort):

Content freshness (update frequency)

Question-based headings

First-60-word answer placement

Priority 2 (more content effort):

Structured data markup

FAQ section implementation

Comparison table creation

Citation, authority data, etc.

The business impact: why this matters now

The early mover advantage

Market timing: We're still in the short window where AI search focus & optimization provides asymmetric returns. Companies optimizing now capture market share before competitors understand the shift.

Customer & trafic quality: AI traffic conversion rates reported up to 3–4x higher

Example of ROI projections based on study data

Hidden Gems companies (strong perception, low visibility) represent the highest ROI opportunity:

Mistertemp': 18% visibility, 67% perception - massive amplification potential

Agryco: 17% visibility, 50% perception - solid foundation for growth

ChapsVision: 12% visibility, 67% perception - technical excellence ready for scaling

The Competitive Moat Effect

Network effects: Companies building comprehensive source ecosystem presence create competitive moats that will be expensive and time-consuming to replicate.

Authority compounding: Early AI citation success creates positive feedback loops, cited companies get cited more frequently, building exponential authority advantages.

Ready to dominate AI search in your industry? Get Started on Mint →

FAQ:

Who conducted this study and why should I trust the results?

This study was conducted by Mint in partnership with Les Echos, France's leading business publication. Les Echos has been covering French business and economic news for over 100 years and is the authoritative source for Next40 analysis. The methodology was reviewed to ensure journalistic standards and objectivity.

How was this study produced?

The study was conducted between June and July 2025. We analyzed 4,500 AI responses across 4 major models (ChatGPT 4o, Gemini Flash 2.0, Mistral Medium, Perplexity Sonar Pro). For each Next40 company, we ran 112 specific queries: 28 queries per model testing both visibility and perception using industry relevant prompts. Over 25,000 unique sources were consulted by the AI models, spanning 3,000 unique domains.

What does the Visibility Score measure?

The Visibility Score represents how often each company was mentioned in AI responses to relevant industry queries. We calculated this as the number of mentions divided by 100 total queries (0-100%). For example, if a company appeared in 95 out of 100 AI responses about their industry, they scored 95% visibility.

What does the Perception Score measure?

The Perception Score analyzes sentiment in AI responses about each company. We asked direct questions about companies (like "What do you think of X?" or "What are the pros and cons of Y?") and categorized responses as positive, neutral, or negative using our analysis system. The score represents: (positive responses - negative responses) / total responses, similar to a Net Promoter Score approach.

What does the Content Score measure?

The Content Score evaluates how well each company's website content is optimized for AI extraction across 11 parameters in 3 categories: Technical (structured data, clean HTML), Content Structure (freshness, headings, meta descriptions), and Content Depth (citing sources, comparison content, in-depth guides). We analyzed 100 random pages per company and scored them on their AI-readiness.

How is the Overall Score calculated?

The Overall Score is our proprietary "Mint Global Score" that combines Visibility, Perception, and Content scores with weighted importance based on their impact on AI search performance. Companies with higher overall scores are more likely to be recommended when customers ask AI for solutions in their industry.

What is GEO (Generative Engine Optimization)?

GEO is the practice of becoming visible in AI-powered search engines like ChatGPT, Gemini, and Perplexity to generate more business through AI recommendations of your products and services. Unlike traditional SEO that focuses on Google rankings, GEO ensures your company appears when customers ask AI for solutions in your industry. Mint is one of the leading and most advanced platform for GEO optimization.

What is Mint?

Mint is a leading platform for Generative Engine Optimization (GEO). We help companies become visible in AI search by providing analysis, optimization tools, and strategic guidance to ensure brands appear when customers ask ChatGPT, Gemini, and other AI models for recommendations in your industry.

Can Mint help me improve my AI visibility?

Yes. Mint provides a complete 3-step approach:

1. Place the sensors:

We test your current brand visibility using calibrated prompts that force AI to cite recommended brands without directly asking (e.g., "What solution do you recommend for managing accounting?"). This critical step requires careful calibration to ask the right questions.

2. Understand the sources:

LLMs rely on sources we classify into three categories:

Your own assets (website, social media) that you control

Influence content (blogs, comparison sites, specialized media) where you can take action

Competitor sources that are out of reach but useful for positioning

This analysis identifies the right action levers.

3. Take action:

Optimize your content on two levels:

Technically (structure, readability, crawlability for LLMs)

Strategically (create or publish content in the right influence zones, with writing designed for AI)

Mint provides tools and guidance for all these steps. Start now for free.

How do I measure ROI from AI search optimization?

Track six metrics: visibility percentage across AI models, perception sentiment scores, brand alignment in responses, source ecosystem influence, content optimization scores, and conversion attribution from AI-driven traffic.

What's the investment required compared to traditional SEO?

AI search optimization requires different expertise and resource allocation compared to traditional SEO. The focus shifts from link building and keyword optimization to content depth development, source ecosystem management, and cross-platform presence optimization.

Should we stop traditional SEO?

Not at all, but rebalance resource allocation. Maintain technical SEO fundamentals while shifting focus and budget toward AI-specific tracking and optimization (esp. content and source ecosystem development).

Which AI platforms should we prioritize?

It depends on your market and product. The key question is: which AI interfaces do your potential customers actually use to discover, compare, and choose solutions like yours?

How do we handle negative AI sentiment?

Implement systematic reputation management across key perception sources: G2.com, Glassdoor, Trustpilot. Monitor sentiment weekly, respond to criticism constructively, and amplify positive mentions.

The Inflection Point: Act Now or Fall Behind

The Window Is Closing

Market timing: We're at the inflection point where AI search optimization shifts from early competitive advantage to business necessity.

Industry transformation: Just as companies that ignored mobile optimization in 2010 or social media in 2012 faced existential challenges, businesses ignoring AI search optimization today risk customer acquisition irrelevance.

Hence:

→ The status quo is not an option. Your customers are already using AI for purchase decisions. The question isn't whether to optimize for AI search, it's whether you'll lead or follow in the transformation.

Transform your AI search performance today Start Your AI Search Transformation →

Study Credits: Les Echos x Mint Exclusive Partnership, July 2025

Methodology: 4,500 AI responses analyzed across 40 Next40 companies and 4 major AI platforms

Source Analysis: 25,000+ unique sources across 3,000 domains

AI is no longer just a trend, it’s a tectonic shift in how customers search, evaluate, and choose solutions.

AI Search Reality Check: How Generative Engines Are Rewriting Customer Acquisition