How Google finds, understands and shows a website

Four steps, in the same order every time#

Google itself describes its search as three steps after an address has been discovered: crawling, indexing and serving [1]. Every page has to pass the step before to get further. Google does not promise that every page is included, not even when everything is done right [1].

• Discovery. Google finds the address through links or the sitemap [2].
• Crawling. Google's program reads the page, if robots.txt allows it [3].
• Indexing. Google understands what the page is about and stores it. Duplicates are merged into one canonical address [4].
• Serving. For every search, Google chooses which stored pages are shown and in which order.

How long does it take? Google's own figures#

On 2 October 2026, Gary Illyes of Google showed timings for 23 different steps at the Search Central Live conference in Barcelona. The figures were written down by John Campbell of the agency ROAST, who was there [5], and repeated by Search Engine Journal [6]. Google has not published the underlying data, so see them as benchmarks, not promises.

How to read the picture: the filled bar is the usual case, the dashed line shows how far the slowest reported case reaches. An arrow at the end means “or never”. That is almost always about quality: Google does not spend time on pages it does not think are worth it. The scale is logarithmic: the distance from an hour to a day is about as long as from a day to a month.

What it means for you#

• A changed title or heading usually shows in the results after 1-2 days, but Google may choose to show a different text than the one you wrote [7] [8].
• A new page is usually found within a day and can be indexed a few hours after that. Whether it then ranks high is another matter: that is decided by the competition and by how helpful the content is [9].
• Big changes in position often happen in connection with Google's core updates, several times a year. The reported recovery time is 3-6 months [5] [10].

This is what we do so that no step stops you#

• A sitemap and a robots.txt that let the search engines in [2] [3].
• A title, a meta description and a main heading (H1) on every page that lacks them. If the heading is already large bold text, we mark it up as an H1 without changing how it looks.
• Structured data about the company (name, address, phone) when it is missing. It must match what is shown on the page, otherwise it breaks Google's rules [11].
• Image texts (alt text) and faster loading of images further down the page.
• When you move to us: the same addresses and the same redirects as today (read more in Switch without losses).

We cannot buy or request a better position, and nobody else can either. Google does not charge for reading a site more often or ranking it higher [1].

Read on#

The tags that control the title and text in the search result are explained in Meta tags, and how the link looks when it is shared in Sharing on social media.