UX glossary · engl. Information Scent
Information Scent
Information scent describes how well people can judge from cues like link text whether a click will lead them to what they are looking for.
Imagine you want to find out on your city council’s website how to renew your ID card. The menu shows “Citizen Services”, “Living in the City” and “Town Hall News”. Any of them could fit, none fits clearly. You click at random, end up on an events page and go back. After the third miss, you give up and search Google instead.
What’s missing here is a good trail.
Explained simply
Information scent describes how well people can judge, based on cues, whether a link or page will lead them to what they’re looking for. The comparison comes from the animal world: an animal follows the scent of food. On websites, people follow cues like link text, headings and images. If the scent gets stronger, they keep going. If it gets weaker, they turn back or give up.
The term comes from information foraging theory, which models how people hunt for information the way animals forage for food. Peter Pirolli and Stuart Card developed it in the 1990s at the PARC research center in Palo Alto. They took models from behavioral ecology that describe how animals weigh effort against reward when foraging and applied them to information seeking. In 1999, they published the theory in the journal Psychological Review.
What makes up the scent
Nielsen Norman Group describes information scent, in essence, as people’s estimate of how much value a source will deliver. That estimate draws on two things. First, what people see: the link text, a short description next to it, images and the surrounding page. Second, what they already know: experience with a brand, domain knowledge or recommendations from others.
One key point: scent is always relative to the goal. A “Plans” link smells strong to someone comparing prices and weak to someone who wants to report an outage. It’s also only an estimate. It can lead people astray, and that’s exactly when frustration and drop-offs happen.
Closely related is information architecture. It decides which terms appear in menus and on overview pages in the first place. You can read more in An Introduction to Information Architecture.
An example
Back to the city council, as an example of a redesign. Before: three vague menu items that promise everything and nothing. After: the team runs a tree test to see which labels people expect their tasks under. One section is now called “ID Cards, Passports and Registration”, with the most common tasks listed below it. The target page opens with the heading “Renew your ID card”. Anyone who clicks sees right away: this is the right place.
In practice
- Write link text that describes the destination. “Learn more” or internal product names give no scent. Words from your audience’s own vocabulary do.
- Confirm the scent on the landing page. The heading and first screen should echo the promise of the link. Otherwise people start to doubt and turn back.
- Skip the clickbait. Sensational teasers bring clicks in the short term. But people who feel misled trust the next links less.
Related terms
Information Architecture, Tree Testing, Card Sorting, UX Writing