How to Show Up in Claude AI: Win on Brave Search First

For the last year, getting your brand mentioned by AI has felt a bit like guesswork. Everyone has got a theory. Half the industry is selling "get cited in ChatGPT" packages. And most of it, hand on heart, is people making confident noises about a black box none of us can fully see inside.

Then some data lands that actually gives you a lever to pull. New analysis shared by Jonathan Clark of Moving Traffic Media, from a Zero Click by Profound session, suggests Claude is far more predictable than the rest. And the practical takeaway is brilliantly simple: if you want to show up in Claude, focus on ranking well in Brave Search.

Let me explain why that matters, because there is real opportunity here for UK brands paying attention.

Claude does not reinvent the wheel, it borrows Brave's

The standout finding is that Claude "does not re-rank search results." When Claude does go and search the web, it appears to lean directly on Brave's top 10 results and use them in its answer.

Think about what that means. Most AI engines feel like a slot machine. You cannot see the logic, you cannot track it, you cannot correlate your effort to an outcome. Claude looks different. If its answers are drawn from Brave's top 10, then Brave rankings become something you can actually monitor and improve. That is a measurable input feeding a measurable output. For anyone who likes their marketing accountable rather than mystical, that is a big deal.

And to be clear, this is early data from one practitioner's testing shared on LinkedIn, not a Google-sized study. So treat it as a strong signal to test, not gospel. But it is the most actionable signal I have seen in months.

Claude is fussier about when it searches

Here is the other half of the story. Claude does not search the web for everything. It only ran a web search on 36.6% of prompts in the testing, compared with around 90% for ChatGPT.

When Claude does not search, it cannot cite a web page. So the game is not just ranking well, it is understanding which questions actually send Claude looking in the first place. The pattern is clear:

• "Best" and "top" style prompts triggered a search 81% of the time
• Ranking-style prompts around 67%
• Location prompts like "near me" around 55%
• Comparison prompts, "X vs Y", around 51%

Meanwhile prompts like "how does", "what is" and "steps to" tended to stay in Claude's own memory, no search, no citation. So a how-to explainer that is perfect for Google might never get a look in from Claude, simply because Claude never goes hunting for it.

For a clothing retailer, that is the difference between "what is merino wool" (Claude answers from memory) and "best merino base layers 2026" (Claude searches, and now your Brave ranking is in play). Same brand, same expertise, completely different visibility depending on the question. Knowing which is which is half the battle.

SEO graft is not wasted

One more finding that should reassure anyone who has invested in proper search work. Claude's citations overlapped with Google's rankings 64% of the time, but only 8% with ChatGPT's.

In plain terms, the strong organic foundations you have built for Google carry over to Claude far more than any ChatGPT-specific trickery does.

That is the bit I love. There is no magic spell. The same honest work that earns you Google rankings, useful content, clean technical SEO, genuine authority, is the work that earns Claude citations too. Google said much the same this week, warning that buying your way into AI citations counts as spam. The shortcuts do not work. The fundamentals do. Same as it ever was.

What this means for your brand

So if you want a slice of Claude's growing audience, here is where I would start:

1. Start tracking your Brave Search rankings. It is the input most UK marketers are completely ignoring, and it now looks like a direct line into Claude. You cannot improve what you do not measure.
2. Map your priority queries by intent. Which of your money keywords are "best", "top", "near me" or "X vs Y"? Those are the ones that trigger a Claude search. Prioritise content and rankings there.
3. Keep doing the Google work. With 64% overlap, your organic foundations are doing double duty. This is not a new channel to build from scratch, it is extra mileage from work you are already doing.
4. Put the year in your titles where it makes sense. Claude's searches often fan out with the current year attached, so "best running shoes 2026" may have an edge over a dateless title for freshness-driven queries.
5. Do not bother buying citations. It does not work, and Google is now calling it spam. Build the real thing instead.

The honest headline is this: Claude may be the most optimisable AI answer engine out there right now, precisely because it behaves consistently and ties back to rankings you can see. While everyone else is chasing ChatGPT shadows, there is a clearer, more measurable path sitting right here.

We are already testing this across client accounts at Clicky. If you want to know whether your brand is showing up when people ask Claude about your category, and what to do if it is not, get in touch. This is exactly the sort of thing we love getting our teeth into. 👊

Clicky™ / ClickyGroup™

Source: Jonathan Clark, Moving Traffic Media, via a Zero Click by Profound session, reported by Search Engine Land. Figures are from early practitioner testing, so worth confirming against your own data before betting big.