The URL Inspection tool says "URL is on Google." The Performance report says zero. Not three impressions, not a bad average position — a flat line. It feels like a contradiction, and it is the point where most people start resubmitting the URL, which changes nothing at all.
It is not a contradiction. The takeaway up front: being indexed makes a page eligible to appear; an impression means it was actually selected and shown for a query somebody typed. Eligibility is a technical state you control. Selection is a competition you enter. Zero impressions means the page has entered that competition and never once been picked — and there are only three real reasons for that, plus a handful of reporting artefacts that fake it.
Work through them in order. Half of all "zero impressions" cases end at the artefacts, before you touch the page.
What an impression actually counts
An impression is recorded when a link to your page appears in a result set that a user loaded. Position ten on page one counts. Position eight on page four counts, if somebody paged that far. It does not require a click, and it does not require a good position.
That definition explains the whole problem. A results page holds a limited number of listings — roughly ten organic results, varying by layout — and almost nobody paginates. So a page sitting at position 70 for every query it might match will show approximately zero impressions, forever, despite being perfectly indexed. The index tells you the page exists in the system. Impressions tell you whether the system ever had a reason to show it.
Keep the ladder straight, the same way we keep submitted ≠ indexed ≠ ranked distinct: indexed → eligible → selected → shown (impression) → clicked. Zero impressions is a failure at "selected," and that stage has nothing to do with sitemaps, pings or submission.
Step 1: rule out the reporting artefacts
Before you rewrite anything, make sure the zero is real. These four cost ten minutes and regularly account for the entire mystery.
Data freshness. The Performance report is not live. The most recent day or two is routinely incomplete or missing, so a page published this week can show nothing simply because the report has not caught up. If your date range ends today, you are partly reading an empty bucket.
The wrong property. Search Console properties are scoped. A URL-prefix property for
https://www.example.com/ does not report on https://example.com/ pages, and a
property added for the http:// version reports on a site that effectively no longer
exists. A domain property covers every subdomain and protocol; if you have several
properties, check the one that actually contains the URL, and prefer the domain property
for diagnosis.
The page filter is stricter than you think. Filtering Performance by page uses an
exact URL. A trailing slash, a different case, ?utm_source= on the version you pasted,
or an /amp/ variant will all return a clean zero while the real URL has data. Filter by
"URL contains" a distinctive path fragment instead of "URL is exactly," and see whether
data appears under a sibling address.
Canonical consolidation. If the engine picked a different URL as canonical for this content, all performance data is attributed to that chosen canonical, not to the URL you are inspecting. URL Inspection shows you the "Google-selected canonical" explicitly. When it differs from the URL you submitted, your page is not missing from search — its numbers are simply filed under another address, and the fix is a canonical/duplication fix, not a visibility one.
Step 2: confirm the page is genuinely indexed
"Indexed" needs to come from the authoritative report, not a vibe check. Search Console's
Pages report and URL Inspection are the two sources that state a verdict and a reason. A
site: search is a crude sanity check that is frequently misleading in both directions.
If the verdict turns out to be "Discovered — currently not indexed" or "Crawled — currently not indexed," you have an indexing problem rather than an impressions problem, and the repair is completely different: our getting indexed guide walks the discovery-crawl-evaluation pipeline and the fixes at each stage. Come back here only when the page is confirmed in the index.
Step 3: the three real causes
With artefacts eliminated, one of three things is true. They look identical in the report and have nothing in common as repairs, so identify which before acting.
| Cause | What the evidence looks like | The actual fix | Realistic effort |
|---|---|---|---|
| No demand — the page targets language nobody types | Site has impressions on other pages; this page's topic returns almost nothing when you search the phrase yourself | Re-target the page at a phrase people search; rewrite title, H1 and opening to match it | Hours — an editing job |
| Not competitive — eligible, but ranked far too deep to be seen | Other pages perform; this one is new, thin, or on a young domain with few links | Add substance, internal links from your strong pages, and a few honest external signals; then wait | Weeks — a compounding job |
| Absorbed by a duplicate — another URL is the one being shown | A similar page of yours has the impressions, or Google-selected canonical differs | Consolidate or differentiate the pages; make one the deliberate target | Hours, then re-evaluation |
No demand is the cause people resist most and it is the most common on new sites. Product pages written in internal vocabulary, service pages named after the brand's own term for a thing, blog posts answering a question phrased the way a founder phrases it — all perfectly indexable, all aimed at strings no member of the public enters. Test it crudely: type the phrase your page is built around into a search engine. If autocomplete does not recognise it and the results look nothing like your page's topic, the demand is not there. No amount of submission fixes a query nobody runs.
Not competitive is the honest state of most new pages on most new sites. The page is in the pile; the pile is deep. What changes it is the same unglamorous set that changes everything else at this stage: a page substantial enough that the engine has a reason to prefer it, internal links from the pages of yours that already get crawled and shown, and a trickle of real external references so the domain stops being an unknown. This is slow by nature and no service can shortcut it — anyone selling "impressions in 48 hours" is selling either bot traffic or fiction.
Absorbed by a duplicate is the quiet one. Two of your own pages covering the same ground, a category page duplicating its top article, or a syndicated copy on a stronger domain — the engine picks one address to represent the content and reports everything there. You do not have an invisible page; you have a page whose data lives elsewhere.
What genuinely moves the number
Ordered by how reliably it works for a page that is indexed and flat, with the reason for the order:
- Retarget the page's language to a real query — highest leverage because it fixes the only cause that time cannot. Put the phrase in the title tag, the H1 and the first paragraph, in the words a searcher uses, not the words your team uses.
- Answer the query more completely than the page currently does — selection is comparative; a page that covers the question end to end gives the engine a reason to surface it over a thinner one.
- Link to it internally from pages that already earn impressions — the cheapest authority you own, and it also tells the engine what the page is about.
- Earn a few legitimate external discovery paths — profile pages, industry registries, and moderated directory listings. Modest and real: a directory listing is a genuine page and a genuine backlink, useful as a discovery and citation signal, not a ranking lever. A listing on a site that accepts anything and is not itself indexed contributes nothing.
- Wait, and measure impressions rather than position — first impressions typically appear on queries you never targeted. That is data, not noise: it tells you what your page is understood to be about.
What does not work, in order of how often it is tried: requesting indexing again on an unchanged page (the request was heard), resubmitting the sitemap daily (it is re-read on its own schedule), and buying bulk submissions or "indexing guarantees" (which promise to overrule a search engine's own judgement about what to show).
When to stop waiting
A new page on an established site that still shows nothing after several weeks of active work is telling you the target phrase is wrong, not that it needs more patience. A new page on a brand-new domain deserves longer — months, honestly — but only if the target phrase survives the crude test above. Patience applied to the wrong cause is just delay.
FAQ
What does "indexed but no impressions" actually mean?
The page is stored in the search index and eligible to be shown, but was never selected for any result set a user loaded. Indexing is a technical state; impressions are the outcome of a comparison against every other eligible page. Being indexed guarantees eligibility only.
How long before a new page gets its first impressions?
There is no fixed timeline and anyone stating one is guessing. Days to a few weeks is common on a site that already gets crawled and shown; a brand-new domain with few inbound links can take considerably longer. Judge by whether impressions have started appearing at all, not by position.
Does zero impressions mean my site has been penalised?
Almost never. Penalties and manual actions are visible in Search Console's Manual Actions and Security Issues reports — check those and you have your answer in a minute. Zero impressions on one page with other pages performing normally is a targeting or competitiveness problem, not a punishment.
Can directory listings or backlinks fix zero impressions?
They help with one of the three causes — competitiveness — and only modestly. Real links from pages that are themselves crawled and indexed create discovery paths and lift a young domain out of obscurity. They cannot create demand for a phrase nobody searches, and links from unmoderated link farms contribute nothing worth having.
Should I request indexing again?
Only after you have genuinely changed the page; then a single request is a reasonable signal that something is new. Repeated requests on an unchanged URL do nothing, and the habit of resubmitting is what pushes people toward spam blasts.
Fix the cause, then send the signals
Diagnose first — artefacts, then index status, then which of the three causes your report actually points at. Once the page targets a real query and says something worth showing, the discovery work is the part worth automating: submit your URL to the directory and an approved listing gives you a public page and a real backlink, while the toolkit fires sitemap and URL pings plus social-share distribution and reports what each submission produced. Honest ranges, no impression or ranking guarantees — from us or anyone.
