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.

Biggest divergence — 75 point gap
Outreach
ChatGPT
95
Claude
20
Gemini
20
Same 5 queries, identical phrasing. A sales leader asking ChatGPT for a recommendation encounters Outreach. The same leader asking Claude or Gemini doesn't.

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:

Second-largest divergence — 55 point gap
Linear
ChatGPT
45
Claude
100
Gemini
95
Linear is widely regarded as one of the best issue trackers in the startup world. Claude and Gemini both recommend it confidently. ChatGPT is oddly quiet about it.

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: The Enterprise Legacy Bias
Overweights established US enterprise sales and marketing tools

ChatGPT scores these brands significantly higher than Claude or Gemini for the same queries:

Outreach +75 vs Claude Salesloft +30 vs Claude Carta +30 vs Claude Otter.ai +25 vs Claude Miro +20 vs Claude Midjourney +20 vs Claude

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: The Startup & Developer Bias
Systematically favours newer, dev-native, and VC-backed tools

Claude scores these brands significantly higher than ChatGPT for identical queries:

Linear +55 vs ChatGPT Render +40 vs ChatGPT Ramp +35 vs ChatGPT Railway +30 vs ChatGPT Close +25 vs ChatGPT Clay +25 vs ChatGPT Remote +25 vs ChatGPT Cursor +20 vs ChatGPT ElevenLabs +20 vs ChatGPT Perplexity +20 vs ChatGPT

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: The Challenger & HR Bias
Favours HR/people-ops tools and some challengers over incumbents

Gemini scores these brands notably higher than ChatGPT:

Linear +50 vs ChatGPT Gong +25 vs ChatGPT Render +25 vs ChatGPT Rippling +20 vs ChatGPT Ramp +20 vs ChatGPT Netlify +20 vs ChatGPT Close +20 vs 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
Outreach95202075CRM
Linear451009555Productivity
Salesloft85553550CRM
Render40806540Dev Tools
Ramp651008535Fintech
Carta100709030Fintech
Railway65957530Dev Tools
Remote50754530Fintech
Close40656025CRM
Gong70859525CRM
Clay55806525CRM
Otter.ai100757525AI Tools
Miro1008010020Productivity
Netlify65808520Dev Tools
Cursor801009520AI Tools
ElevenLabs801009520AI Tools
Perplexity801009020AI Tools
Midjourney95759020AI Tools
Rippling75809520Fintech
Descript1001008020AI 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.

Figma
100 · 100 · 100
Vercel
100 · 100 · 100
Postman
100 · 100 · 100
Supabase
100 · 100 · 100
Sentry
100 · 100 · 100
Gusto
100 · 100 · 100
HubSpot
95 · 100 · 100
Stripe
95 · 100 · 100
Plaid
95 · 100 · 100
Airtable
95 · 100 · 95
Mercury
90 · 100 · 100
Notion
95 · 90 · 100
PlanetScale
90 · 100 · 95
Jasper
100 · 95 · 90
Deel
90 · 95 · 100
Retool
85 · 95 · 95
Brex
90 · 90 · 95
Salesforce
85 · 95 · 85
Asana
90 · 80 · 90
Pipedrive
70 · 80 · 75
Coda
70 · 75 · 65
Attio
65 · 70 · 60
Runway
55 · 65 · 60

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:

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 →

Frequently Asked Questions

Do ChatGPT, Claude, and Gemini recommend the same software?
Not always. For well-established tools, all three platforms largely agree. But our research found significant divergences — particularly in CRM and sales software, where some brands score 75+ points higher on one platform than another for identical queries. The AI engine you use can determine whether a brand appears in your shortlist at all.
Why does ChatGPT recommend different software than Claude?
Each AI engine was trained on different datasets and weights different content sources differently. ChatGPT appears to overweight traditional enterprise sales publications, making it more likely to surface established US sales tools. Claude appears to weight developer communities and startup-adjacent content more heavily, making it more likely to recommend newer, dev-native tools.
Which AI gives the most balanced software recommendations?
Based on our data, Claude scored brands highest on average (88 vs ChatGPT's 83 and Gemini's 85), suggesting it is the most generous recommender. However "balanced" depends on context — Claude's startup bias makes it excellent for discovering newer tools, while ChatGPT's recommendations skew toward established enterprise software.
What does this mean for software vendors?
It means your AI visibility strategy cannot be platform-agnostic. A brand that scores highly on ChatGPT may be nearly invisible on Claude or Gemini. Software vendors need to audit their visibility across all three platforms separately and understand which content types and sources influence each platform's recommendations.