The report
Your report scores your brand across ChatGPT, Claude, Grok and Perplexity, shows the sources those answers are built on, and turns it into concrete steps — each with a brief and a prompt you can paste to draft the work.
01
Three numbers, 0–100: how often models name you, whether they cite your own site, and how much of your category’s citations you hold. Each one comes with a short explanation of why it is what it is. The scoring is identical every run, so you can compare later.
02
The ranked list of sites models actually cite when someone asks a buying question in your category: competitors, review sites, docs, a forum thread or two. Usually not who you’d guess.
03
A short list, ranked by impact. Each item tells you what to do, which score it moves, and what “done” looks like — and comes with a prompt that drafts the work. Some are new pages. Some are a rewrite, a directory listing, or one well-placed reply.
How it works
That’s the only input. We read the site, understand the product, and generate high-intent category prompts — no questionnaire.
Grok, Claude, OpenAI, and Perplexity answer the same buyer questions. We extract citations and brand mentions from those answers.
A report with the three scores, the sites getting cited, and the ranked list of what to do — each step with a brief and a prompt. It takes a few minutes to run, and we email you the moment it’s ready. Yours to keep and share; there is no account to log into.
From a real run
A revenue-analytics company, anonymised. Models know them — but only when you name them first. Ask which tool to use and they don’t come up at all. These are the four things their report told them to do, each with a brief and a prompt to do it.
/guides/how-to-choose-revenue-analytics
Buyers ask “how do I pick one” constantly. Right now models answer with a competitor’s blog post from 2023.
/vs/[category leader]
The leader will never list you on their comparison page. This is the page models can cite instead.
Want to read a whole one first? Open the full sample report — same report you’d receive, brand anonymised.
Who it’s for
Buy it if
Skip it if
Pricing
The audit now, and a re-run 45 days later once you’ve shipped — same scores, same prompts, so you can see exactly what moved. One payment, no subscription.
$95 USD
Launch price · going to $120
Both runs are included. We’ll email you when it’s time for the second one — there is nothing to cancel and nothing to renew.
Progress loop
A score you can’t compare is trivia. Your re-run uses the same scores and the same prompts, so the difference is your work — not a new opinion. It’s included in the price, and we email you when it’s time.
Ship the pages in the report. Drafts and internal docs do not change organic visibility, brand as source, or graph presence.
Give crawlers and models time to find the new URLs. Re-running the next day would look identical — that’s why we wait.
Same three scores, side by side with the first run, plus whether the pages you shipped are actually getting cited. No extra charge and nothing to request.
FAQ
Grok, Claude, OpenAI, and Perplexity. Same high-intent prompts across models so the source map is comparable, not a single-vendor anecdote.
Fair question, and the one most tools dodge. Every score is an aggregate across sixteen buyer prompts and four models — sixty-four answers, never a single sample — and the report shows the raw counts behind each number so you can see how solid it is. When a re-run delta is small enough to sit inside normal answer-to-answer noise, the report says so instead of selling it to you as progress.
Yes, and some people should. Open four chats, ask twenty buying questions, log every citation, dedupe by domain — that’s an afternoon. The afternoon isn’t the hard part. Deciding which two or three moves are worth the week, and what has to be true for each to get cited, is. That decision is what you’re buying.
No. We do not track Google positions. CitePeek measures whether your brand gets named and cited inside the answers ChatGPT, Claude, Grok and Perplexity give your buyers — and which other sites those answers trust.
A few minutes. The run starts the moment your payment clears, and we email you the moment the report is ready.
The website URL and an email to send the report to. That’s the whole intake — we read the site, infer the category and the competitor set, and build the prompts ourselves. No questionnaire, no kickoff call. Raw prompts and raw model responses stay out of the report; every finding drawn from them does not.
Yes — one re-run, 45 days after your first report, at no extra cost. Ship the top actions, and we re-measure the same three scores with the same prompts and show both runs side by side. We email you when it’s due; you don’t have to remember or ask.
No guarantees. CitePeek shows where you stand and which moves are most likely to make you a citable source. Models choose citations; we do not sell rankings in any assistant.
Those products are built for ongoing tracking across many prompts and engines. CitePeek is closer to a Lighthouse audit than a dashboard: one run, one clear picture of who owns citations in your category, and a short ranked list of what to publish or change — each with a brief and a prompt. If you want something that updates every week, use a monitor. If you want a prioritised execution list from one serious run, use this. Plenty of teams should have both.
One report for one website, plus its re-run 45 days later — $95 total, paid once. Auditing a second site is a second report. $95 is the launch price; it goes to $120 after launch, and whatever you paid covers both of your runs.
No. You pay once and get two runs: the audit now, and the re-run 45 days later. There is no monthly plan on purpose — the product is the plan, not the login.
Get the map
Send a URL, get back the ranked list of what to change and the briefs to do it. $95 for both runs while it’s the launch price — the first one is ready in minutes.
Get your report — $95