In the November 2026 edition of the CiteHawk AI Index, the nine AI models we track named three or more different number-one brands, with no clear majority, in 121 of 255 categories. They agreed on a single number one in just 39.
So if you're checking one AI model to see where your brand stands, you're seeing one answer out of several. Depending on the model, its number one differs from the overall number one in a third to a half of all categories.
Every number below comes from the frozen November 2026 edition. Each category was asked the same buying question across ChatGPT, Claude, Gemini, Perplexity, DeepSeek, Grok, Copilot, Google AI and Google AI Mode, and the full answers are archived on each category's edition page.
How often the AI models agree
We sorted every category by how its models' top picks fell:
| What the models did | Categories | Share |
|---|---|---|
| All named the same number one | 39 | 15% |
| At least three in four named the same number one | 80 | 31% |
| Split two ways, no three-in-four majority | 15 | 6% |
| Split three or more ways, no three-in-four majority | 121 | 47% |
Put another way, 216 of the 255 categories had at least two different number ones. In 68 of them, the most popular pick didn't have even half the models behind it.
Where they fall apart
The widest splits are worth looking at, because they show how far apart the answers can get.
Ask for the best marketing conference and you get eight different number ones from nine models: Content Marketing World from Perplexity and Google AI, then Social Media Marketing World, Adobe Summit, BrightonSEO, Cannes Lions, the Content Marketing Institute, MarketingProfs and SXSW, one from each of the others.
Managed IT service providers split seven ways. Claude and Google AI Mode named Accenture, ChatGPT and Google AI named TeamLogic IT, and the rest went to IBM, Integris, Kyndryl, Electric.ai and Capgemini. Coding bootcamps split seven ways too, from App Academy (Perplexity and Copilot) and Codesmith (Gemini and DeepSeek) to five other schools.
None of these are fringe questions. They're the kind of thing a buyer types into whichever AI model they happen to use, and the answer depends on which one that is.
Where they agree
The 39 unanimous categories lean heavily toward software with one dominant player: accounting software, password managers, video conferencing, code hosting, data warehouses, CDN providers and ERP software among them. A handful of consumer and service categories made the list as well, including food delivery apps, car rental companies, meal kits, mortgage lenders and personal injury lawyers.
When one brand owns a category this clearly, the models don't need to weigh much up. Everywhere else, they do, and each one weighs things its own way.
Which models sit closest to the consensus
For each model, we measured how often its own number one matched the overall number one for the category (the brand that ranks first once all nine models' answers are combined):
| AI model | Matches the overall number one |
|---|---|
| Claude | 69% |
| Perplexity | 68% |
| Google AI Mode | 67% |
| Grok | 65% |
| Gemini | 64% |
| ChatGPT | 62% |
| Google AI | 62% |
| DeepSeek | 58% |
| Copilot | 48% |
Copilot is the outlier. Fewer than half of its top picks matched the overall number one, so a brand that checks only Copilot is the most likely to get a misleading read on where it stands.
What it means for your brand
If your team checks one AI model, it's measuring one view of the market. In nearly half of categories, the other models are putting someone else first.
Three things follow from that:
- Check every model that matters to your buyers. A strong result on one model tells you very little about the others.
- Look at where you're named, not just whether you're first. In a split category, being the second or third pick on several models can be worth more than being first on one.
- Track it monthly. In November, 62 categories got a new overall number one compared with October. These rankings move.
You can see the full November picture, category by category, in the November 2026 Index Report. If you want to see where your own brand sits, run a free report.
Rankings in the CiteHawk AI Index are computed from AI answers alone. They're never for sale.
