What 83 pages with zero impressions had in common

Search Console has two ways of telling you a page is doing nothing, and on screen they look almost identical. Zero clicks means Google put your page in front of people and nobody picked it. Zero impressions means Google never put it in front of anybody at all. Same empty row, completely different repair job — and most of the advice you will be given applies to the first one.

I know which one I have, because I counted mine. My site has 359 published pages. For the six months to 11 September 2026, Google recorded zero impressions for 83 of them — 23% of everything I have ever published here. So I went through all 83 and asked Search Console, one page at a time, what it thought was wrong with them.

The answer was: nothing. Seventy of the ones that are supposed to be findable are fully indexed. Google has read them, stored them, and never once shown them to a human being.

What does zero impressions in Search Console mean?

It means that during the dates you selected, Google never displayed a link to that page on a results page somebody actually looked at. That is the whole definition. It is not a penalty, not a manual action, and not a warning. It is the absence of an event, and the event can be missing for reasons that have nothing to do with your page.

Everything below is one measurement of one website: leadlls.com, 359 pages, Search Console data for 15 March to 11 September 2026, pulled read-only on 14 September 2026. Index statuses come from the URL Inspection API, which reads a status and cannot submit anything. Where I quote Google, the exact page is linked at the bottom.

Zero impressions and zero clicks are not one problem

Before you change anything, find out which row you are actually looking at. A page with 900 impressions and no clicks has been offered to 900 people who read your title and kept scrolling — that is a writing and position problem. A page with no impressions has never been offered to anybody, so the title cannot be blamed for anything.

If it is the first row you are in, the useful reading is what "average position" is actually an average of, because that number is what decides whether anyone ever sees the title you are about to rewrite. The rest of this is about the second row.

Step one: subtract the pages that are supposed to be invisible

Five of my 83 were never meant to be found. The thank-you page someone lands on after sending a form, the cancelled-payment page, the invoice payment page, and two archived versions of legal documents kept for the record. All five carry a noindex tag that I put there on purpose. They are not failures; they are correctly behaving pages sitting in a report that does not know what they are for.

Do this first, because it is free and it changes the shape of the list. Sort your zero-impression URLs and cross off everything that exists to be arrived at rather than found: confirmations, receipts, internal tools, archived terms, print versions. Whatever is left is the actual question.

Step two: ask Google whether it even has the page

This is the ten-second check almost nobody runs before they start rewriting. Paste the URL into the search bar at the top of Search Console and read the status it gives back. There are only three answers that matter, and they send you in three completely different directions.

Ninety per cent of my supposedly broken pages are not broken. And here is the part that finished the argument for me: I ran the same inspection on 40 pages that Google does show, spread evenly across the site. All 40 came back "Submitted and indexed" too. The status is identical on both sides. It cannot be the explanation for either.

Indexed means Google kept a copy. It has never meant Google intends to use it.

The two pages that embarrass me most are both in the 70. My guide on getting found on Google is 1,434 words, was last crawled on 10 August, is fully indexed — and has been shown to nobody in six months. Same for my website maintenance checklist. If a technical fault were causing this, those two would have it, and they do not.

It is not thin content, and I can prove it on my own site

The standard answer at this point is "the page is too thin". On my site the opposite is true, and it is not close. The median zero-impression page runs 1,791 words. The median page Google actually shows runs 1,127. My silent pages are 664 words longer than my successful ones.

I am not going to claim that length hurts, because that table has an obvious confound and I can see it: my newest pages are also my longest ones, and new pages have had less time to be shown. What the table does kill is the reverse claim. There is no reading of this data in which the silent pages are the short ones. My longest page on the entire site — 5,882 words, on pest control web design — has zero impressions. My shortest real page, a 235-word link preview checker, gets shown.

Internal links do not explain it either. I counted every internal link on this site last week — 24,250 of them — and 71 of the 83 silent pages already had one pointing at them from inside a sentence. I am not going to re-run that argument here.

What it actually is: the phrase, not the page

My site has two large sets of pages built by scripts from fixed templates. One set targets a city — "web design Mississauga". The other targets a trade — "web design for funeral homes". Same code, same author, same design, and the median number of inbound links written into a sentence is three in both sets. The city pages first appeared in Search Console in the first half of July; the trade pages in the second half. Two weeks apart, and that is the only age difference between them.

Forty per cent more trade pages, earning one tenth as many impressions. And the query list says why. Across the six months Google reported 1,630 distinct queries for my property. 808 of them name a place, and those carry 81% of every impression I received. Queries shaped like "web design for something" — the exact shape 123 of my pages were written for — account for 91 queries and 547 impressions in total, which is 1.6%.

Then there is the blunt version. Fourteen words never appear in a single one of those 1,630 queries, in any spelling: dealership, funeral, travel, veterinary, retirement, masonry, soccer, catering, deck, fence, pest control, mortgage, appliance, pool. I have 29 pages built around those words. Between them, in six months, they collected 15 impressions — and since no query contains any of the words, not one of those fifteen came from somebody looking for that trade.

I built 123 of those pages around one phrase shape — “web design for X” — which turns out to carry 1.6% of my search demand. The city pages use the shape that carries 81% of it. Nothing else about the two sets differs, and one earns ten times the other. Publishing more pages does not create the demand; it only spreads the same demand thinner.

The honest limit: Search Console cannot tell you which of the two it is

Here is where I have to be careful, because the test I just ran is weaker than it looks. "No query containing the word funeral" does not prove nobody searches for it. It proves that Google never placed my site anywhere a person looking for that word actually looked. Search Console only reports queries that produced an impression, so a phrase you rank 40th for looks exactly like a phrase nobody types.

And the numbers refuse to separate cleanly. Thirty of my 39 silent trade pages have no query carrying their trade word — but so do 42 of the 84 trade pages that are getting shown. The test leans. It does not decide.

Google's own rule is what makes the two indistinguishable. An impression is counted, in their words, "whenever an item appears in the current page of results, whether or not the item is scrolled into view, as long as the user need not click to see more results". And on position: "A link must get an impression for its position to be recorded. If a result does not get an impression — for example, if the result is on page 3 of search results, but the user only views page 1 — then its position is not recorded for that query."

Read that twice. If you sit on page three of a busy phrase and nobody clicks through to page three, Search Console reports zero impressions and no position — the same two zeros you get for a phrase nobody searches at all. The report is not hiding anything from you. In both cases there is genuinely nothing to record.

The ten-minute check that does separate them

Since the dashboard cannot answer it, answer it from outside. This takes about ten minutes for five pages and it costs nothing.

1. Paste the URL into Search Console's top search bar and read the status. If it is not "Submitted and indexed", stop here — you have an indexing question, not a demand question, and it is a different job.
2. Open a private browser window and type the exact phrase the page targets into Google. Not your brand name. The phrase in your H1.
3. Look at what came back on page one. Ten strong, on-topic pages from real competitors means the demand is there and you are below where anyone looks. Google returning loosely related pages, forum threads and dictionary definitions means almost nobody is searching this and the page will never be shown regardless of what you do to it.
4. Check whether one of your own pages is already ranking for it. Add site:yourdomain.com to the same search. If a different page of yours comes back, you have two pages fighting over one phrase, and the older one is usually winning.
5. Decide in one line. Demand exists and you are buried: the page needs to be genuinely better than what is on page one, or it needs a narrower phrase. No demand: leave the page alone and stop spending on it.

Two things that will confuse you while you look

Your query numbers will not add up to your total. My page report says 48,997 impressions for the six months. My query report says 34,303. That is a 30% gap and it is not a bug: Google withholds rare queries, describing them as omitted "to protect user privacy". You cannot get those back, and you should not build a theory on the difference.

Pages you deleted are still earning. Twenty-nine URLs that no longer exist on my site collected 1,939 impressions in the same window — twenty-two retired project pages and seven old service URLs. If you compare your sitemap against your Pages report, those will show up as strangers, and they are real data rather than a mistake in your spreadsheet.

If your Search Console is full of zeros and you want somebody to tell you which of the three statuses you are actually in, send me the domain — that is a twenty-minute answer, not a project. If the pages are indexed and the problem is that they sit below where anyone looks, that is SEO work and it starts with the phrases, not the pages. If the site itself is the reason nobody stays, that is a rebuild. What I charge is on the pricing page, and the trades I actually work in are on the industries page — including several I have now proven nobody searches for.

Sources, read in full on 14 September 2026: Google — Search Console performance data for the verbatim definitions of an impression and of position, including the page-three example; Google — Performance report dimensions and data groupings for anonymized queries. Everything numeric here is my own measurement rather than an industry study: page and query figures are Search Console data for leadlls.com from 15 March to 11 September 2026, joined to the 359 index.html files on disk by URL path; word counts were parsed from the same files on 14 September 2026; coverage states come from the Search Console URL Inspection API, which is a read — no URL was submitted for indexing, and every failed read was recorded as unknown rather than counted as "not indexed".