Google search snippet preview

The honest answer is "about 600 pixels", not a number of characters. Google cuts by width, with its own font, and ten "i" take up what four "M" do. The 60-character rule works as an approximation because most titles mix wide and narrow letters; a title in capitals gets cut much earlier and one full of "l" and "i" lasts longer. That is why the tool measures in pixels and marks the cut where it really happens.

Because it thinks another text describes the page better for that particular search. It does so mostly when the title does not match the H1, when it is a string of keywords separated by bars, when it repeats across many pages of the site, or when it is too long and Google would rather rewrite than cut. A title that says the same as the H1, in one sentence, with the brand at the end, is the one rewritten least often.

Not directly: Google has said many times it is not used as a ranking signal. But it decides the click. Between two results in the same position, the one that explains better what is inside gets more visits, and that does end up mattering. And if you do not write one, Google replaces it with a passage of the page it chooses, sometimes a piece of the menu or the footer.

It is the tag that says "of all the addresses this page can be reached at, the official one is this". Without it, the same page with and without a trailing slash, with and without www, or with a campaign parameter, counts as several different pages sharing the authority. A wrong canonical is worse than none: if it points at another address, Google indexes that one and drops yours without warning.

Google knows which version to show someone searching in Spanish and someone searching in English, but not someone searching in German. x-default is the version for "everyone else", usually the main language or a selector. Also, every page has to include itself in its hreflang group and the others have to link back: a one-way hreflang is ignored entirely.

Yes, when you type a URL: the browser cannot read another domain on its own, so a server of ours requests the page and passes your browser only the beginning of the document, up to </head> or 256 KB at most. Nothing is stored; the result is cached for ten minutes so the same page is not requested twice while you fix things. All the analysis happens afterwards, in your browser. If the page sits behind a password, paste the code by hand and it never leaves your machine.

Yes. Paste the page <head> — or the whole document — into the text box and it is analysed the same way. It is how to check titles and descriptions before a new site goes live, while everything can still be changed at no cost.