Information Scent

It comes from information foraging theory, developed by Peter Pirolli and Stuart Card at Xerox PARC. They modelled web users as foraging animals: following the trail that smells most likely to lead to food — here, the information they're after — and abandoning trails that go cold.

Users navigate by following scent, not by systematic exploration. Pirolli and Card's information foraging theory — developed at Xerox PARC and now one of the foundational models of web behaviour — shows that users behave like foraging animals: they follow the path that seems most likely to lead to their goal, and abandon it the moment it stops smelling productive. The scent of a link is the user's estimate of how likely that destination is to satisfy their current need, derived from the link label, surrounding content, images, and prior experience. Vague, jargon-heavy, or misleading labels don't just confuse users — they cause them to miss content that would have served them perfectly, because they never clicked the link. Strong information scent means every label and navigation option accurately and concretely describes what the user will find.

The 'scent' of a link is the user's estimate, from its label and surroundings, of how likely the destination is to satisfy their need. People follow the strongest-smelling trail and abandon paths the moment they stop seeming productive — so a weak or misleading label doesn't just confuse, it makes useful content go unfound.

This is why vague and jargon-heavy labels are so costly. A link the user can't read scent from is a link they may never click, even when what's behind it is exactly what they wanted. Strong scent means every label concretely describes what the user will actually find.

Information scent is the user's sense, from the cues available, of how likely a given path is to lead to what they want. It comes from information foraging theory — Pirolli and Card's model of web users as foragers who follow the trail that smells most promising and abandon ones that go cold. The scent of a link is built from its label, its surrounding content, nearby images, and the user's prior experience.

Users navigate by scent, not by systematic exploration. They scan, pick the option that seems most likely to lead to their goal, and follow it only as long as it keeps smelling productive. The moment a path stops seeming relevant, they back out and try another — or leave. Navigation is a series of scent judgements, made fast and largely unconsciously.

The practical lesson is that labels carry enormous weight. A vague, jargon-heavy, or misleading label doesn't just confuse — it causes users to miss content that would have served them, because the weak scent meant they never clicked. Strong information scent means every label and link concretely and honestly describes its destination. It connects to satisficing, message match, and self-evident design.

Kweri can assess some aspects of information scent directly — whether navigation and link labels are concrete and descriptive or vague and jargon-heavy, and whether the path to important content is clearly signposted. What it can't fully judge is whether a label's scent matches what your specific users are looking for, since that depends on their vocabulary and intent. So Kweri flags weak, ambiguous, or generic labels that likely give poor scent, and prompts you toward concrete descriptions, while confirming the scent leads users to their goal is a matter of testing.

Identified by Peter Pirolli and Stuart Card. Catalogued from Nielsen Norman Group — Information Scent.

From Pirolli and Card's information foraging theory, developed at Xerox PARC; the linked NN/G article is the reference used here.

Read the primary source →