What is a user agent? The user agent string explained

What is a user agent?

A user agent is whatever software makes a web request on someone's behalf: a browser, a mobile app, a search engine crawler, an AI crawler, a monitoring script. It introduces itself in the User-Agent header.

A sample string, decoded

Mozilla/5.0 (iPhone; CPU iPhone OS 18_0 like Mac OS X) AppleWebKit/605.1.15
(KHTML, like Gecko) Version/18.0 Mobile/15E148 Safari/604.1

A crawler identifies itself differently, for example with a name such as Googlebot or GPTBot and a link to its documentation.

What analytics reads from it

• Browser, operating system and device type for audience reports.
• Device brand and model, where the string includes them (common on Android).
• Declared bots and crawlers, which are separated from human traffic (bot, AI crawler).

Reduction and client hints

Long strings made browsers easy to tell apart, so Chrome now sends a reduced user agent: minor version numbers are frozen and platform details trimmed. Sites that really need the full version or device model can ask for User-Agent Client Hints (Sec-CH-UA headers). Expect less detail from the plain string over time.

Limits

• It can be faked. A scraper can claim to be Chrome on Windows; a fake crawler can claim to be Googlebot. Verify crawlers by IP or reverse DNS and look at behaviour.
• It is not a person. Many people share the same string, and one person uses several.

In TapCub, device brand and model in the audience report come from the user agent, and the cookieless visitor ID uses the user agent with version numbers removed, so a browser update does not create a new visitor.