The audience you actually have
Find the pages ranking for the wrong intent.
QueryScope groups every search Google names for your site by the job behind it, sizes each group against the rest, and tells your agent to decide which of them is your buyer before you spend an afternoon on any one page.
From $49/mo, everything included · works in Claude Code, Cursor, Cline and Windsurf
Built from the Search Console audit we run on our own SaaS, AppScreenshotStudio.
Why it hides
The traffic grows, and it belongs to somebody else's buyer.
A site can pull thousands of impressions a month from people who will never buy anything, and every number a dashboard shows will look like progress: clicks up, position holding, impressions climbing. Volume cannot say who those visitors were or what they came to do. Only the searches can, and only in aggregate: one page at a time tells you nothing, because the audience that is quietly wrong is spread across every page you own. How a page ends up ranking for the wrong intent.
The totals that look fine
A green number that cannot say who it came from.
The same rows, grouped by the person
The audience you have. Whether it is yours needs the repo.
Two cuts of your own search data, and one boundary that keeps the verdict where it belongs.
Grouped by the job behind the search, not its subject.
Five jobs, named in the words the searcher would use: comparing tools or checking a price, wants it free, learning how to do it themselves, looking up a number or a spec, looking for a tool to use. Each row carries its searches, impressions, clicks, share and examples, plus one line on what it is worth commercially. Comparing is the closest thing to a buyer that exists in search data, because the person has named a vendor or asked a price. A job needs at least three searches and a hundred impressions before it is named at all, and what carries no clear signal is reported as its own share rather than folded into a guess.
wants it free · 1,090 impressions · 10% of everything shown
How a page ends up ranking for the wrong intentThe same rows, cut by subject, with the one keeper named.
A different axis on purpose: the job cut hides a subject, and the subject cut hides a job. This one clusters the searches you appear for by what they are about, gives each cluster its share of your whole search footprint and the share of it carrying buying intent, and names the single buying-intent search inside an otherwise research-only cluster along with the position you already hold on it. A cluster like that is more than wasted effort: it is what Google reads the site as being about.
25% of your footprint · 4% of it shopping · one search worth keeping
Reading a real account this way for 30 daysIt names and sizes. Your agent, holding the repo, decides.
Whether an audience is your buyer needs to know what the site sells and who it sells against, and that lives in your repository rather than in Search Console. So QueryScope groups and measures and stops there, and the agent with your code open answers the question. Record what the site is once with the brief and the read aims at it from then on. Two honest bounds ride with the numbers: the shares are shares of the searches Google names for you, since rare ones are anonymised out, and both sides of every share come from that same pool, so the percentage holds even where the raw counts run low.
grouped and sized by us · judged by the agent with the repo
Page-level vs query-level dataOn a live site
The plan board, on a SaaS we run ourselves.
AppScreenshotStudio's declared searches, read from production: where each one stands and whether it moved.
How the plan is doing
21 May to 25 Sep · arrows are places moved, up is better
On page 1
where clicks happen · was 0 on 21 May
Below page 1
ranking, earning almost nothing · was 2 on 21 May
Not ranking
no reading on 10+ impressions · was 55 on 21 May
Coverage
<1%The whole site earned 528 clicks in 28 days, most of it on searches this plan never claimed.
On page 1
figma app store screenshot template
#9.9
▲3
320/mo
Demand
38
Impressions
2
Clicks
One line in your IDE. Then authorize and pick a site.
claude mcp add --transport http queryscope https://mcp.queryscope.dev/mcpWorks the same in Cursor, Cline and Windsurf. Your IDE opens a browser once to authorize.
Then ask: “what should I go after?”
FAQ
Reading intent in your own data.
How does QueryScope work out that I am pulling the wrong audience?
It groups every search Google names for your site by the job behind it (comparing a price, wanting it free, learning to do it themselves, looking up a number, shopping for a tool), sizes each group against your whole footprint, and prints one line on what each is worth commercially. Then it says the part it cannot answer out loud: whether a job is your buyer needs to know what you sell, which is in your repo rather than in Search Console. So the numbers come from us and the verdict comes from your agent.
Which tool does this, and when do I call it?
The keyword tool with no arguments, which is the first call on any site. It reads back what is stored for the site and leads that answer with the audience you already have, because ranking higher inside the wrong job pours in more of the traffic that never converts. It deliberately sits outside the working loop: work hands back one search you declared at a time, and the page pulling the wrong audience is by definition a search nobody declared.
Is this a rank tracker or a keyword tool?
No. It reads your own Google Search Console data over an OAuth grant you approve and can revoke, so the searches are the ones your pages already appear for, not a third-party estimate. It groups demand you already have rather than inventing new terms to chase.
What counts as a search that will never pay?
Searches that read as buying and rarely do: the free-seekers (free, gratis, freeware) and the people hunting a pirated build (apk, crack, keygen, torrent). Both fold into one row, 'wants it free', because the commercial read is the same. Price shoppers (cheap, cheapest, discount, coupon) are deliberately left as buyers: that person has already chosen the category and is filtering on price, which converts.
Will fixing the mismatch make more money?
QueryScope cannot say. It reads your search data and stops at the click. What a visitor does after they arrive is a different measurement, in your analytics or your product. It shows you who is arriving and what they came to do; the fix and its payoff are yours to measure.
See who is finding you, and what they came to do.
Connect your site, tell QueryScope what you sell and who you sell against, and let your agent read the audience before it picks a page to work.
7-day free trial · nothing due today · cancel in one click
Other things it finds in your Search Console