Glossary
Title tag
The `<title>` element gives a document its name. The browser shows it in the tab, a bookmark takes it as a default label, and a search engine uses it as the primary source for the clickable link in a result. It is required in every HTML document, there must be exactly one, and it is the single piece of text most likely to be read by someone who has not yet decided whether to visit the page.
Where the title is used
The browser tab and window, where it is often truncated to a few characters and where a distinctive first word matters.
Bookmarks, history and the text a link takes when someone copies a page into a document or a chat.
The result link in search. Google generates that link primarily from the title element, and rewrites it when it judges another source describes the page better.
Sharing previews, when no Open Graph title is present — Open Graph takes precedence where it exists.
Why search engines rewrite titles
Google documents that it may replace a title link with text drawn from headings, link text or other on-page content, when the title is a poor fit.
Common triggers: a title that is empty or boilerplate, one that repeats a site name many times, one stuffed with keywords, or one that does not match the page's actual content.
Rewriting also happens when the title is very long and truncation would cut it mid-idea.
The remedy is not a longer or shorter title but an accurate one: a title that describes the page is rarely replaced.
What makes a title work
It describes this page, not the site. Every page having the same title is the most common serious title problem, and it makes a set of results indistinguishable.
The distinctive part comes first. Truncation removes the end, and a title beginning with the site name spends the visible part on something the reader already knows.
It matches the page. A title promising something the page does not deliver costs more in bounces than it gains in clicks.
There is no character limit in the standard. Search results truncate at a width rather than a count, which is why the guidance is to front-load rather than to hit a number.
How to check a site's titles
`curl -s https://example.com/page | grep -o '<title>[^<]*</title>'` reads one page's title from the served HTML.
Across a site, the questions worth asking are: is any title missing, is any duplicated, and is any so long that its distinctive part is beyond the truncation point.
Duplicates matter more than length. Two pages with the same title are two pages a search engine has to disambiguate without help.
For pages rendered client-side, check the rendered DOM as well: a title set by JavaScript is used, but it is not in the source.
Frequently asked questions
- How long should a title tag be?
- There is no defined limit. Search results truncate by pixel width, so the practical advice is to put the distinctive words first rather than to target a character count.
- Why does Google show a different title than mine?
- Google rewrites titles it judges unhelpful — boilerplate, keyword-stuffed, or mismatched with the page. An accurate, specific title is rarely replaced.
- Can a page have two title elements?
- No. HTML permits exactly one, and a second is a markup error with unpredictable handling.
Sources
Related
VeriFixScan crawls a site and applies its checks to every page it reaches, keeping the evidence behind each finding. Scanning one website is free.
Scan a website