TrafficTriageSearch Console Impressions vs Page Views: Why They Differ
By Ugur Saritepe · October 28, 2026
Search Console impressions and analytics page views are not two counts of the same visits. On game-scout.app, 28 days to 2026-10-14, Google reported 145,660 impressions and GA4 recorded 6,581 page views. The ratio between them ran from under 1 to 60 depending on the kind of page, and most of the pages with page views had no impressions at all.
Where this came from
game-scout.app is my other site, a catalogue of curated game lists. In September I was reading its numbers in three dashboards and typed this to the assistant I work with:
the numbers change a lot between vercel, ga4 and gsc actually.
That session ended up about clicks against GA4 sessions, which I wrote up as why GA4 shows fewer visits than Search Console clicks. The comparison I had not run was the one people ask about most: impressions against page views, page by page. I ran it on 2026-10-17 over 2026-09-17 to 2026-10-14, the last 28 days Search Console had complete.
The headline numbers
sc-domain:game-scout.app, 2026-09-17 to 2026-10-14, totals
Search Console impressions 145,660 clicks 2,377 position 8.42
GA4 page views 6,581 sessions 3,377 (same dates)
22 impressions per page view, site-wideTwenty-two to one sounds like a single fact about the site. It is not. Split by the kind of page, the same 28 days give very different answers:
page type impressions page views impressions per page view /best lists 117,181 3,760 31 /articles 27,689 462 60 /game pages 747 1,905 0.4 home page 318 290 1.1 /search 1 104 0.01
Lists and articles are what Google shows. Game pages and search are where people go once they are on the site. The site-wide 22 is an average of two different things: how often Google puts a link in front of someone, and how much browsing happens after the first page.
What each number counts
Google's help page on impressions says an impression is counted when a link to your page is in the current page of results, whether or not it is scrolled into view. Scrolling away and back in the same search does not count twice. Inside a carousel or an expandable box the link usually has to be scrolled to or opened. So an impression means your link was on a results page that loaded, not that anyone read it, and only on Google.
A page view is the other end. GA4 counts one each time a page of the site loads with the tag running, whatever brought the visitor there: Google, Bing, a bookmark, or a click from another page on the site. On game-scout the 3,377 sessions made 6,581 page views, about two pages a visit. The second page of a visit has no impression behind it.
Most pages are on one side only
I joined the two by page path, with www and non-www folded together. GA4 recorded views on 692 paths and Search Console reported 421. Only 253 had both.
pages page views impressions clicks page views, no impressions 439 1,680 0 0 impressions, no page views 168 0 5,144 11 both 253 4,901 140,973 2,366
439 of the 692 pages people viewed, 63%, had no Search Console row in 28 days. They carry 26% of the page views, and 1,409 of those are on game pages, which Google showed 747 times in the whole window. One of them is the PC flight games list, 81 page views, which I found the day before is Bing's best page on the site and a URL Google has never heard of. Its visitors came from somewhere other than Google, and Search Console will never show it.
The other side is smaller and stranger. 168 pages had impressions and not one page view. The biggest is a comparison of two Monster Hunter games: 3,230 impressions at an average position of 7.9, no clicks, no page views. That is a page Google shows on page 1 and nobody picks, which is what happened to most of our comparison pages.
The ratio per page, and why it moves
On the 58 pages with at least 500 impressions and at least one page view, impressions per page view ran from 6 to 417, with a median of about 45. A few of them:
page impressions clicks page views ratio position best-cleaning-chore-games 2,759 191 187 15 6.1 best-trucking-games 8,440 165 278 30 6.9 best-flight-games 12,519 171 176 71 8.6 best-ps-plus-racing-games 3,630 14 31 117 8.2 It Takes Two co-op guide 5,837 4 14 417 8.8
Position explains some of it. Pages averaging 5 to 8 had 31 impressions per page view, pages at 8 to 10 had 45. It does not explain the co-op guide, which sits at 8.8 like the flight list and has six times the ratio: 5,837 impressions and 4 clicks. I have not read the results pages it shows on, so I cannot say what takes the click instead. What the table does show is that two pages at the same position can sit an order of magnitude apart, so the ratio is not a score you can read off position alone.
On the pages averaging position 5 to 10, page views ran 1.5 to 1.8 per Search Console click. Even where Google sends the visit, the page view count includes the people who arrived some other way, and impressions without clicks is its own diagnosis.
The page views are an undercount
One more reason the ratio is not a fact about the site: our GA4 misses visits. The site sets Consent Mode to deny analytics storage by default in 32 countries and has no banner to change it, so GA4 records about 18 sessions for every 100 Google clicks from those countries, measured in the GA4 against Search Console case. Every ratio above is against the page views GA4 recorded, not the page views that happened. If your analytics drops visitors too, your impressions will look even further ahead.
What the ranked pages say
I checked three ways of asking this on 2026-10-17. Google's own help page is in the top three for all of them. The rest are definition pages from analytics and marketing vendors, forum threads, and pages about social media impressions against video views. They explain what an impression is, correctly. None of the pages I read puts one site's impressions next to its page views, by page.
What we have not done
We have not shipped the consent banner that would let GA4 count those visitors, so the page view side of this comparison stays low in Europe. The third number in my September line, Vercel's page views, still cannot be read: its analytics API returned "Web Analytics not found" for the project again on 2026-10-17. And TrafficTriage reads Search Console, not GA4: the page view half of this join came from a script.
How to compare them in ten minutes
- Use the same dates on both, ending three days ago. Search Console runs two to three days behind, so the last days of a "last 28 days" view are not complete yet.
- Export impressions by page and page views by page, then join on the path. Fold www and non-www, drop trailing slashes and query strings.
- Split the result into three groups: both, page views only, impressions only. Page views only are pages people reach by browsing or from other engines. Impressions only are pages Google shows that nobody picks.
- Read the ratio only for pages in both, and next to position. Good: a list at position 6 with 15 impressions per page view. Worth a look: a page on page 1 with a ratio ten times its neighbours, where the results page may be answering the question for you.
- Check your analytics for missing visitors before you trust the page view side. Consent defaults and bot filters both lower it.
If the impressions-only group is mostly pages you expected people to visit, the free triage report checks whether their titles and snippets are the problem. The other measured reads in this series are in AI-assistant SEO.
FAQ
Are Search Console impressions the same as page views?
No. An impression is counted when a link to your page is on a Google results page someone loaded, whether or not they scrolled to it. A page view is counted when your page loads in a browser, from any source. On game-scout.app over 28 days that was 145,660 impressions against 6,581 GA4 page views, and 63% of the pages with page views had no impressions at all.
Why are my impressions so much higher than my page views?
Because most impressions never become a visit. On our site the pages ranking between position 8 and 10 had 45 impressions for every page view, and one comparison page had 3,230 impressions, no clicks and no page views. The ratio depends on position and on what else is on the results page, so there is no normal number to aim for.
Can a page have more page views than impressions?
Yes, and on most sites many will. Our game pages had 747 impressions and 1,905 page views in 28 days, and almost none of those visits came from a Google result. Page views count internal navigation, other search engines and direct visits. Impressions count only Google results.
Which number should I use to judge SEO?
Use impressions and clicks from Search Console to judge how Google shows you, and page views to judge what people do once they arrive. Compare them page by page, not as site totals, and only on pages that have both. A site total of impressions per page view mixes pages Google shows nobody with pages nobody reaches from Google.
The impressions half of this join is one tool call. Connect Search Console to the TrafficTriage MCP server and ask your assistant for impressions and clicks by page over the dates of your analytics export. Each call returns up to 250 pages with exact totals, and the join happens in the reply. It does not read GA4; the page views you still bring yourself.
In your AI assistant
Ask your AI about your own site
TrafficTriage plugs into Claude, Cursor, and ChatGPT. Your assistant runs the 8 checks, reads your Search Console data, and can even check your site every morning. Free account, no API key, nothing to install.