GenGrowthObservation
Impressions, clicks and average position for every search query with enough volume to measure. Queries, not pages: that is the only dimension this report reads.
Search performance evidence
Find SEO Opportunities in Google Search Console
Turn high-impression, low-click queries and positions 8–20 into a prioritized action list.
Requires a Google Search Console connection
GenGrowth requests read-only Search Console access. It cannot publish pages, change rankings, or modify your Google account, and it stores nothing on our servers.
Connect Search Console
No demo data, nothing stored on our servers. Every number is computed from the property you choose.
Connect Search Console
We read your Search Console performance data — impressions, clicks, and average position, query by query. Read-only: we cannot change anything on your site or in your account. We do not store your results on our servers. The authorization itself is kept encrypted in a cookie in your own browser, so a return visit does not start from the consent screen again; it ends the moment you disconnect — here on this page, or from your Google account.
Connect Search Console
Free · Read-only access · Nothing stored on our servers · Disconnect anytime
How to find high impressions with low clicks
- 1
Connect Search Console
Read-only, one click, revoke it whenever you like from your Google account.
- 2
We build your site's own CTR curve first
What counts as a normal click-through rate on your site, at each position band, computed from your own non-brand queries over the last 28 complete days.
- 3
Start with what to do next
Every run opens with a short list of moves. Each one is gated on a number in the same report and names the rows it came from — including the moves not to make. None of them predicts what a change would do.
- 4
Then read the gaps, largest first
Every search query whose click-through rate falls below what your own site earns at that position, ranked by the size of the shortfall. Each row says which checking path it is on, so you can filter to the handful worth a morning. Sort any column, or take the whole table away as CSV.
- 5
Where we can, we show you a query's page next to one that does better
Drawn from your own site and named, so you can open it and judge the comparison yourself. No comparable page, no draft.
What you get, and what it stops short of
Your own baseline
What a normal click-through rate looks like on your site at each position band, with the query being measured excluded from its own baseline.
The gap
Expected clicks minus observed clicks. A measured difference between two observed numbers, not a forecast of what a change would recover.
What to do next
A short list of moves drawn from this run: which rows need a look at the search results page, which conclusions the evidence does not support, and where the picture is incomplete enough that you should go elsewhere for it. Every entry names what put it there.
What we cannot explain
We do not detect AI Overviews or other SERP features, and those are a leading cause of these gaps. The tool names that limit on every run.
What one finding looks like
- Observation
- The query “lamine yamal zodiac sign”, average position 8.9, 3,439 impressions, 3 clicks — a click-through rate of 0.09% over 28 days.
- Compared against your own site
- Queries on this site at positions 8–11 earn 0.51% on average across 451 queries and 16,885 impressions. With this query removed from its own baseline, the rest of its band earns 0.62%. So the site's own curve predicts about 21 clicks for those 3,439 impressions, and 3 arrived: a gap of about 18 clicks over the same 28 days.
- What we do not claim
- We cannot tell you why. A gap this size can come from a title that does not match the query, from searchers wanting something this page was never meant to answer, or from Google answering the question inside the results page so nobody needs to click. Those need different fixes, and one of them is not fixable at all.
- What the report tells you to do about it
- Not to rewrite this one on its own. Its whole band earns 0.51%, so every query in there falls short of baseline for the same structural reason, and treating this row as its own problem turns one fact into 451. The report says so in as many words and points at the band rather than the title. It also sends this property to the Pages report, because 445 of its 451 queries never cleared the impression floor — per-query is the wrong resolution for a site whose demand is spread that thin.
- What you can take away
- A short list of what to do next, each entry naming the rows that put it there, over the full evidence table — every query with its observed rate, your site's rate at that position, the gap between them, the tail probability, and which checking path the row is on — sortable by any column and exportable as CSV.
What this finds, and how it ranks it
High impressions, barely any clicks
Any search query with at least 100 impressions, sitting in a position band your site has enough data to measure, whose click-through rate falls below what your own site earns in that band. The 100 is a flat floor, not a share of your traffic: below it, one click either way moves the rate enough to change the answer.
Improve organic CTR without touching rankings
When the cause is how your result is written, the feedback loop is days rather than months. When it is not, no rewrite will help — which is why this page shows you the evidence rather than a verdict.
Measured against your own site, not an industry table
Published CTR benchmarks say roughly 28% at position 1 and 2% at position 10. We tested those on a real site and found positions 4–10 earning a tenth of that, while 11–16 earned three times what 4–10 did. Industry averages describe a plain blue-link results page. Your site may not have one. So we build the curve from your data.
Easy SEO wins, ranked by the size of the gap
Sorted by what your site's own curve predicts minus the clicks observed, largest shortfall first. Rows that beat your own curve are kept deliberately, with a negative gap. That is a measured difference between two observed numbers, not a forecast of what you would recover.
A draft you can see the source of
Where your own site has a comparable page earning a clearly higher click-through rate at a similar position, we name that page and draft a title on the same pattern. No comparable page, no draft — we do not fall back to a generic template. At most five drafts per run, on the largest gaps.
The whole table, out of the tab
Nothing is stored on our servers, so the export is the copy that survives. It carries the same rows, the same measured window and the same blanks — a value we could not compute stays an empty cell rather than becoming a zero.
Who this is for
Sites that already rank for something. If you have queries landing in positions 4–20 with real impressions behind them, there is usually a measurable gap sitting in them. If most of your pages have no impressions at all, this is not your problem yet — start with SEO Agent.
The query that made us build this — and what it taught us
One of our own sites had a query sitting at position 8.9, pulling 3,439 impressions over 28 days and converting three of them into clicks. A 0.09% click-through rate. Nothing was broken: the page ranked, it was indexed, it loaded fine. No audit tool we ran flagged it, because by every technical measure the page was healthy.
We spent an evening on it and concluded the title was weak. We were wrong. The query was “lamine yamal zodiac sign” — and Google answers that inside the results page. Nobody needed to click. No title rewrite was ever going to move it.
That is why this tool shows you the gap and names what it cannot explain, instead of handing you a diagnosis. Finding the gap is the part software does well. Deciding what it means still needs you to read a few of the actual queries.
How this compares
Google Search Console for beginners — where this fits
Search Console gives you the raw numbers. It will not tell you which queries are underperforming relative to what your own site achieves at that position, or how big each gap is. That comparison is the whole job of this tool.
How this differs from Traffic Drop Diagnosis
Same underlying data, different question. Use this one to find gaps in what you already have. Use Traffic Drop Diagnosis when something has already fallen and you need to know when and by how much.
How we decide a click-through rate is too low
Your site's curve, built from your own 28 days
We group your non-brand queries by average position and compute the impression-weighted click-through rate for each band. A query is compared against its own band on your site, and it is excluded from the band it is measured against — so nothing can set its own benchmark.
When we cannot build the curve
A position band needs at least 500 impressions and 5 distinct queries before it can serve as a baseline. Bands below that produce no findings, and we tell you which ones and why. On small sites this can mean most bands — we would rather say so than compute a number from four queries.
What counts as enough impressions to measure
A flat floor of 100 impressions per query. These thresholds are absolute, not relative to your site. On a small site that honestly leaves single digits, and we list which queries were excluded and for which reason rather than padding the table.
What this will not tell you
Why the click-through rate is low
This is the big one. A gap can mean your title is weak, or that searchers wanted something else, or that Google answered the query in the results page. We measure the gap; we do not detect which. Read a few of the actual queries before rewriting anything.
Whether an AI Overview or featured snippet is taking the clicks
We do not detect SERP features. Search Console does not expose them in a way we can line up with the rest of this data, and we would rather leave the question open than guess.
Anything about queries Google hides
Search Console withholds low-volume queries for privacy. On the site we tested, 46% of impressions and 64% of clicks were not in the query report at all. Everything here is computed on what Google returns, which is not everything — and every run tells you how large that missing share was for your property.
Which page a query belongs to
The report reads one dimension, queries, so a row is a search term rather than a URL. Grouping a query by page is a second, less complete read that Google drops rows from, so that split is not part of this table.
What happened after you changed something
We do not store your results, so there is nothing to come back to. Export your current numbers before you edit, then re-run this in a few weeks and compare — Search Console keeps the history even though we do not.
Related tools
- SEO Agent
Reviews public HTML without Search Console; account verification is required before the crawl.
- Tech Agent
Reviews crawl, indexability, and internal-link evidence without saving the run.
- Traffic Drop Diagnosis
When traffic has already fallen.
Related reading
- From 0 to 5,000 users: the astrologywiki.com case study
The site the numbers on this page come from.
- What a public SEO audit can and cannot see
Where first-party data stops being optional.
FAQ
What does high impressions with low clicks mean?
Google is showing you in search results and people are choosing not to click. Impressions count how often you appeared; clicks count how often you were chosen. A wide gap between the two means you are visible but not being picked.
Is a low CTR always a problem?
No. A 1% click-through rate at position 18 is ordinary. The same 1% at position 3 usually is not. What matters is your rate relative to what your own site achieves at that position, which is the only comparison this tool makes.
What is a normal CTR for my position?
Published benchmarks say roughly 28% at position 1, 11% at position 3 and 2% at position 10. Treat those as a rough map rather than a measurement. We tested them against a real site and found positions 4-10 earning about a tenth of the published figure, because that site's queries were being answered inside the results page. Your own curve is the only one that describes your site.
Why do I have impressions but no clicks at all?
Three causes are common: a title that does not match the query, an AI Overview or featured snippet answering above you, or a page ranking for a query it was never meant to serve. This tool finds the queries. Telling the three apart still takes a look at what people actually searched for.
Does this work without Search Console?
No. It reads your own site's private search performance data, which only Search Console has. For a public-HTML review that needs no Search Console connection, use SEO Agent; a verified GenGrowth account is required before its crawl.
What access do you need?
Read-only Search Console access. We cannot modify your site, your account or your data, and you can revoke access at any time from your Google account settings.
Do you store my data?
No. We read Search Console, compute the report, and send it to your browser. Nothing is written to a database, so export what you want to keep before you close the tab. The one thing that persists is the Google authorization itself, encrypted in a cookie in your own browser so a return visit does not start from the consent screen; revoking access in your Google account settings ends it.
Can I keep the results after I close the tab?
Yes, as CSV. The file carries the same rows, the same measured window and the same blanks: a value we could not compute stays empty rather than becoming a zero. It is the only copy that survives the tab.
Where do the title drafts come from?
From your own site. We look for a page in the same position band earning a clearly higher click-through rate, show you which page that is, and draft on the same pattern. If there is no comparable page there is no draft — we do not fall back to a generic template.
It will not tell me why the CTR is low, so what do I actually do?
It tells you where to look, and in what order. A row short by more than a click sends you to the search results page for that query — an AI Overview or a featured snippet takes the click before your title is ever read, we do not observe those, and you can in about a minute. When a whole position band earns under 1%, the report says so and tells you not to rewrite a dozen titles over what is one fact. And when most of your queries are too small to measure at all, it points you at the Search Console report that does see them.
How often should I check?
Monthly is enough for most sites. Title and meta changes take a few weeks to appear in Search Console data, so checking more often mostly shows noise.