When we audited 50 tech brands across ChatGPT, Claude, and Gemini, we expected the results to broadly agree. They didn't. For a significant subset of brands, the AI engine you query determines whether a brand appears in recommendations at all — and the gap can be as large as 75 points for identical queries.
This post digs into the divergences: which brands are caught in them, what patterns explain them, and what it means if your brand is on the wrong side of one.
Data source: This analysis is drawn from our 50 Tech Brands AI Visibility Study, where we ran 250 standardised prompts across ChatGPT, Claude, and Gemini. Each brand received 5 queries: two category, one competitive, one buyer intent, one branded — scored 0–100 on each platform.
The Biggest Divergence in the Dataset: Outreach
Outreach, the sales engagement platform, produced the starkest platform divergence we recorded. Ask ChatGPT for "the best sales engagement platform for startups" and Outreach appears prominently — scoring 95. Ask Claude or Gemini the identical question and Outreach is nearly absent, scoring just 20 on both.
Outreach isn't a niche product — it's one of the most established names in B2B sales software, with thousands of customers and significant media coverage. The divergence isn't about obscurity. It's about where that coverage lives and how it weighted in different training datasets.
The other large divergences tell a similar story from the opposite direction:
Three Bias Patterns in the Data
After mapping all divergences of 20+ points, three patterns emerge clearly. They appear to reflect the different content sources that influenced each platform's training data.
ChatGPT scores these brands significantly higher than Claude or Gemini for the same queries:
These are primarily established US companies with long histories of coverage in mainstream tech publications, LinkedIn content, and enterprise software review sites. ChatGPT appears to have particularly strong training data from these sources — generating confident recommendations for brands that other platforms are more agnostic about.
Claude scores these brands significantly higher than ChatGPT for identical queries:
The pattern is striking: Linear, Railway, Render, Ramp, Cursor, and ElevenLabs are all tools that became prominent through developer Twitter/X, Hacker News, and the indie-hacker/startup ecosystem. These communities generate a specific tone and density of content that Anthropic appears to have weighted heavily — resulting in Claude being notably more likely to recommend tools that are beloved in dev circles but haven't yet dominated mainstream enterprise coverage.
Gemini scores these brands notably higher than ChatGPT:
Gemini's pattern is the least pronounced of the three but shows a mild tendency to rate challenger brands and HR/payroll tools (Rippling, Ramp) above what ChatGPT assigns them. Gemini also shares Claude's enthusiasm for Linear — both platforms give it 95+, while ChatGPT only scores it 45. Gemini's profile appears to sit between ChatGPT's enterprise bias and Claude's startup bias.
All Divergences of 20+ Points
Here is every brand where any two platforms diverged by 20 or more points, sorted by maximum gap:
| Brand | ChatGPT | Claude | Gemini | Max gap | Category |
|---|---|---|---|---|---|
| Outreach | 95 | 20 | 20 | 75 | CRM |
| Linear | 45 | 100 | 95 | 55 | Productivity |
| Salesloft | 85 | 55 | 35 | 50 | CRM |
| Render | 40 | 80 | 65 | 40 | Dev Tools |
| Ramp | 65 | 100 | 85 | 35 | Fintech |
| Carta | 100 | 70 | 90 | 30 | Fintech |
| Railway | 65 | 95 | 75 | 30 | Dev Tools |
| Remote | 50 | 75 | 45 | 30 | Fintech |
| Close | 40 | 65 | 60 | 25 | CRM |
| Gong | 70 | 85 | 95 | 25 | CRM |
| Clay | 55 | 80 | 65 | 25 | CRM |
| Otter.ai | 100 | 75 | 75 | 25 | AI Tools |
| Miro | 100 | 80 | 100 | 20 | Productivity |
| Netlify | 65 | 80 | 85 | 20 | Dev Tools |
| Cursor | 80 | 100 | 95 | 20 | AI Tools |
| ElevenLabs | 80 | 100 | 95 | 20 | AI Tools |
| Perplexity | 80 | 100 | 90 | 20 | AI Tools |
| Midjourney | 95 | 75 | 90 | 20 | AI Tools |
| Rippling | 75 | 80 | 95 | 20 | Fintech |
| Descript | 100 | 100 | 80 | 20 | AI Tools |
The 23 Brands Where All Three Agree
By contrast, 23 of the 50 brands showed near-total platform consensus — all three AI engines scoring within 10 points of each other. These are the brands with what we call AI search consensus: they appear in recommendations regardless of which AI your buyer uses.
The consensus club is dominated by dev tools and fintech infrastructure — categories where brands have clear, well-established identities and dense, consistent coverage across every major content source.
Format: ChatGPT · Claude · Gemini
What This Means for Buyers
If you're using AI engines to research software, the divergences in this data have a direct practical implication: the AI you ask is not a neutral oracle. It reflects the particular slice of the internet it was trained on. A sales leader who exclusively uses ChatGPT for research will encounter a different shortlist than one who uses Claude.
The practical recommendation: for any significant software purchase, run your query on at least two AI platforms and compare. Brands that appear prominently on both are likely the consensus market leaders. Brands that appear on only one platform are worth investigating — they may be well-regarded in a specific community that one AI engine knows better than others.
The "two out of three" trap: 20 brands in our dataset had max divergences of 20+ points. A brand on the wrong side of a divergence doesn't lose visibility with buyers who use one specific AI — they lose it entirely with that segment. If your brand is an Outreach (great on ChatGPT, invisible on Claude), you're invisible to every buyer who prefers Claude. If you're a Linear (invisible on ChatGPT), you miss everyone who defaults to it.
What This Means for Software Vendors
The divergence data reveals a risk that most marketing teams haven't accounted for. Your AI visibility strategy cannot be built assuming all three platforms will treat you the same. Specifically:
- Audit all three platforms separately. An overall score hides divergences. A brand with an average of 65 could be scoring 95 on one platform and 35 on another — two completely different stories for different segments of your market.
- Understand where your coverage lives. If you score highly on ChatGPT but not Claude, your content is probably dominated by traditional enterprise publications. To improve on Claude, you likely need more presence in developer communities, indie-hacker forums, and startup-adjacent media.
- The goal is consensus, not just one platform. The brands in the consensus club above are winning regardless of which AI their buyers use. That's the defensible position. Platform-specific visibility is real but fragile.
Check your platform divergences
Run an AI visibility audit and see your score on ChatGPT, Claude, and Gemini separately — not just as an average.
Run Your Audit →