Glossary

User Agent

A user agent is a small but powerful piece of information included in every browser request. It tells a server what kind of device, operating system, and browser a visitor is using. This allows websites and apps to tailor their content, adjust layouts, or deliver optimized experiences based on the user’s environment.

Because user agents reveal technical details about a device, they also play a key role in analytics, attribution, fraud detection, and performance troubleshooting across digital products.

What Is a User Agent

A user agent is a text string that a browser or app sends to a server whenever it loads a page or resource. It describes the client device in a standardized, machine readable format. Servers read this string to understand how to deliver the best possible response.

A typical user agent looks like this for a Mac using Chrome:

Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_6) AppleWebKit 537.36 Chrome 64.0.3282.186 Safari 537.36

The string contains the device type, operating system version, browser engine, and browser version. This combination gives servers a clear snapshot of the environment requesting the content.



Why User Agents Matter

User agents are used across many parts of the digital ecosystem. Their main benefits include the following.



Helping servers adapt content

Websites can deliver the correct layout or features depending on the device. A mobile browser might receive a lighter or rearranged version of a page, while a desktop user gets a fuller interface.



Supporting measurement and attribution

Attribution platforms often use user agents as part of a probabilistic matching model. When deterministic identifiers are limited, user agents help narrow down which device clicked on an ad and later installed an app.



Detecting bots and suspicious activity

Because bots often reuse generic or unusual user agent strings, they can be flagged. Fraud prevention tools look for patterns that do not match real world browsers or devices.



Debugging and technical diagnostics

Developers rely on user agents to identify browser quirks, rendering bugs, and compatibility issues.



User Agents and Mobile Attribution

In environments where device identifiers are restricted or unavailable, attribution models rely on short lived signals such as:

IP address

Timestamp

Referrer

This combination helps estimate the likelihood that a click and an install came from the same user. This technique is known as probabilistic attribution.

Because user agents change frequently with browser or device updates, they cannot be used for long term tracking and are unsuitable for persistent identification. They are most effective in short matching windows where accuracy remains reasonably high.



Limitations of User Agents

While useful, user agents also come with important constraints.

They can be spoofed

Fraud actors can fake user agent strings to impersonate real devices.

They are not stable identifiers

Browsers update constantly, which means the string may change with new versions.

They reveal limited details

Modern privacy standards encourage browsers to reduce or randomize user agent information to protect individuals.

They cannot reliably identify a unique person

User agents describe a device environment, not the individual using it.



FAQs

What information does a user agent contain

It usually includes the device type, operating system, browser engine, browser version, and rendering engine.

Can a user agent identify a specific person

No. It can help estimate device characteristics but cannot uniquely identify a person.

Do user agents change over time

Yes. Browser updates, OS updates, and device changes alter the user agent string.

Why do attribution platforms look at user agents

They help match pre install activity with post install events during short windows where deterministic identifiers may not exist.

Can user agents be faked

Yes. Bots and fraud tools often modify user agents to disguise their identity, which is why additional fraud detection signals are necessary.



Related Terms