How long Google actually took to index 218 of my new pages

Every client asks me the same thing within a week of a site going live: how long until Google indexes a new page? Usually with a slightly panicked edge, because they published on Monday, searched their own headline on Wednesday, found nothing, and decided the website is broken. For years I answered with the industry's favourite non-answer — "a few days to a few weeks, it depends." This month I stopped guessing and went and measured it.

Between 12 July and 5 September 2026 I published 218 pages across four sites I look after: this one, a stucco contractor in Ontario, an art gallery's journal, and a claims-software blog. I know the exact day every single one went live, because the publishing is logged as it happens rather than reconstructed afterwards. So I pulled the daily Search Console history for all 218 URLs and asked one question of each: what is the first day a human being was shown this page?

The median answer is two days. The rest of what I found was considerably less comfortable.

How long does it take Google to index a new page?

Two days, at the median. Of the 218 pages, 159 have appeared in a Google results page at least once. Half of those did it within two days of going live, 85% within a week, and 99% of the ones that ever appeared appeared inside 30 days. The slowest single page took 34 days. If a month has passed and nothing has happened, waiting longer is rarely what fixes it.

Here is the whole distribution, counted cumulatively. Each row includes everything above it.

The number I trust most here is not the median itself but the fact that it repeated. Three of the four sites came in at a median of exactly two days and the fourth at three — and these sites have nothing in common. Different countries, different industries, different ages, one of them a brand-new domain. Whatever is going on, it is not a quirk of one website's setup.

One honest caveat about the method. Search Console counts an impression when, in Google's words, "a user has seen (or potentially seen) a link to your site in Search, Discover, or News". So the first impression is the first time the page was actually served to somebody, which is a slightly later and much more useful milestone than "the day Google filed it away". It is also a lower bound: a page can be perfectly indexed and never be served once, which turns out to be the whole story.

Why is my new page indexed but still not showing up?

Because indexing and appearing are different events, and only one of them is under your control. On this site 32 pages have never had a single impression. I ran Google's URL Inspection on the 21 of them that are at least two weeks old — old enough that silence means something. Nineteen came back "Submitted and indexed" with a verdict of PASS. Google has them. Nobody searched them.

This is the part I would rather not publish, so let me be specific about it. My guide to Core Web Vitals has been indexed since 18 August, is 49 days old, and has been shown to a human being exactly zero times. The one on building a site that ranks and converts: 45 days, same result. Both are, I think, genuinely good pages. Both answer a question that, in this market, almost nobody types into Google.

The distinction matters because the two diagnoses have opposite fixes. "Not indexed" is a crawling or quality problem — you go and look at robots.txt, canonicals, thin content, internal links. "Indexed and silent" is a demand problem, and no amount of technical work touches it. You wrote a good answer to a question that isn't being asked. That is a research mistake, not an engineering one, and I have made it nineteen times this summer.

Google's definition of the other state is worth having in front of you, because it is quoted wrongly everywhere: "The page was found by Google, but not crawled yet. Typically, Google wanted to crawl the URL but this was expected to overload the site; therefore Google rescheduled the crawl." That is a capacity decision about your server, not a verdict on your writing. Two pages out of 218.

Can you make Google index a page faster?

No, and Google is unusually blunt about it in its own documentation. Three sentences from the page on asking Google to recrawl, which is the closest thing to a speed-up button that exists: "Crawling can take anywhere from a few days to a few weeks." Then: "Requesting a crawl does not guarantee that inclusion in search results will happen instantly or even at all." And finally, for anyone who has been mashing the button: "requesting a recrawl multiple times for the same URL won't get it crawled any faster."

So request it once if it makes you feel better — it is free and it does no harm. Requesting it every morning for a week is a ritual, not a lever. What genuinely moves the needle is duller: the URL being in your sitemap on the day it goes live, and being linked from a page Google already visits often. My 218 pages all had both, which is very likely why the median is two days rather than two weeks. It is also why I stopped submitting URLs by hand entirely — the machinery did the job better than the button.

How long until somebody actually clicks it?

Much longer, and for most pages, never. Nineteen of the 218 pages have ever received a single click — 9%. Where a first click did arrive, the median wait was 22 days after publishing, roughly three weeks after the page first appeared. Across all 218 pages: 5,896 impressions, 52 clicks. A click-through rate of 0.88%.

That number is the honest answer to "when will this page start bringing me customers", and it is nothing like the answer to "when will it be indexed". Indexing is a two-day formality. Being chosen is a different game, played against everyone already on that results page.

Splitting the four sites apart is where it gets useful, because the speed of appearing is nearly identical everywhere and the outcome is not remotely.

Same publishing machine. Same schema, same internal linking, same person writing. Every page on the stucco site has been shown to somebody, and nearly half have been clicked. Every one of those queries has a handful of decent answers and a contractor who wants one. Meanwhile my own web-design pages are competing with every agency in this city on a results page that is mostly ads and a map — which, incidentally, is the entire argument for taking the map pack seriously rather than out-writing thirty competitors.

If you sell something specific in a place, this is very good news. The reason your competitors' blogs look like they're working is rarely that they write better. It is that they picked queries with less traffic and far fewer rivals — which is exactly what the SEO work I do spends most of its time on, well before anybody writes a word.

What actually decides whether a new page gets seen

Five things, in the order they bite. None of them is "wait longer", and only the first two are technical.

The fourth row is the one people miss. Impressions going up with clicks staying at zero is not "it's still warming up" — it means Google is showing you somewhere nobody reaches. On this site, 96% of the pages that get impressions have never been clicked. That is not a patience problem.

What I do now, in order

The measurement changed my own process, so here is the version I actually follow.

And the one thing I stopped doing: refreshing Search Console daily for the first week. The data lags about two days anyway, so what you are staring at on Tuesday is Sunday's silence.

Frequently asked questions

Sources, all read in full on 7 September 2026: Google — Ask Google to recrawl your URLs · Google — Page Indexing report · Google — Search Console performance report metrics. The 218-page dataset is my own: every page my publishing agents put live between 12 July and 5 September 2026 across four sites, matched against their daily Search Console history for 8 January to 5 September 2026, plus a read-only URL Inspection run on 7 September 2026. Two URLs were dropped from the sample because they had impressions predating their log entry, which means they were rebuilds of older pages rather than new ones.