Users draw on link labels, summaries, images, and surrounding content, along with what they already know about a source, to estimate whether it can answer their question. They then consider how long finding an answer might take before deciding whether to click or move on. Information scent is a user’s estimate of a source’s value; it may be inaccurate and does not guarantee content quality. The same source may appeal differently to people with different needs. This overview of the selection process draws primarily on Nielsen Norman Group’s (NN/g) 2020 explanation of information foraging theory.
Start with the label: Does it look like the answer I need?
A link’s label is an important cue before users open a page. For example, someone looking for cheese might consider “Food” worth clicking, while someone looking for facial cleanser would not. The difference is not whether the label itself is good or bad, but whether it matches the user’s current need.
A 2023 NN/g article notes that many people simply ignore links with vague labels. Only a relatively small number of users, usually those who are more motivated, are willing to give them a try. Labels should therefore help people anticipate the content, rather than merely urge them to act:
- “Learn more” does not explain what users will see and offers too little information on its own.
- Even accurate technical terminology may cause the intended audience to miss a useful source if they do not understand it.
- Clear labels also help users rule out irrelevant pages; their purpose is not just to increase clicks.
Next, look at summaries and images: Do the surrounding cues help?
Users may scan a summary beside a link to check the content’s main topic and details. Images should represent the page or category rather than serve merely as decoration. If a summary only repeats a confusing label, it gives users no additional information to judge by.
A historical example in NN/g’s 2020 article illustrates this: someone looking for Christmas table settings is more likely to see a link labeled “Christmas” as useful when it appears beside an image of tableware. In other words, the link has stronger information scent. If it is surrounded by fruit and gift baskets instead, it seems less relevant to that need. However, small screens may not display all the surrounding content, and rapid scrolling may cause people to miss cues, so link labels should still be as specific as possible.
Prior knowledge changes how users understand the same entry point
Familiarity with a brand, a positive previous experience, or a recommendation from someone else can all affect how users interpret a source and whether they are willing to click a poorly described link. Domain knowledge also changes their judgment. In the same article, NN/g gives an example involving university admissions data: someone familiar with the dataset’s name and its location on the website may know to look in a finance-related section, while a newcomer may understand neither the dataset’s name nor the entry point’s label.
We therefore cannot assume that first-time visitors will find answers simply because people familiar with the website can. Recommendations from others can provide additional cues, but they do not mean that a source will necessarily answer everyone’s question.
After clicking, users still need to decide whether to continue
A click is merely a choice based on expectations; users only find out whether the content is useful by reading or using it. If the destination page does not provide enough context early on to show how the content relates to their needs, users may leave without scrolling further, even if relevant information appears farther down. Information scent therefore affects not only which source users choose, but also whether they continue exploring after arriving. An appealing entry point cannot replace content that addresses their needs.
At Procrustes, we turn the questions our clients genuinely care about into articles, with ongoing planning, writing, and updates.
Information scent and source selection: perspectives from across the web
How do cues left by other users affect source selection?
People can use cues about source quality left by others to judge which information is worth examining further. Social foraging theory treats this collective search for information as collaboration within a network, which may be an organized research community or an informal group of contributors or reviewers. These cues can supplement information scent, but users still need to consider whether a source answers their own question.
Does a link’s position affect whether users choose it?
It can. In addition to its label and content, where a link appears on a page can affect how users judge it. NN/g notes that even clearly described links in the right sidebar may be mistaken for advertisements because users are accustomed to seeing ads there, making those links seem less relevant. This is a possibility, not a claim that all links in the right sidebar will be ignored.
How can websites with many categories make it easier for users to choose a relevant entry point?
Websites with many categories can consider using vertical navigation on the left side of the desktop layout. This arrangement can accommodate more top-level navigation items and display specific categories directly, so users do not have to open a broad, generic category before seeing options they can judge for relevance. This is a design choice for desktop layouts with many categories, not a recommendation that all websites or small screens use the same layout.
References
- Left-Side Vertical Navigation on Desktop: Scalable, Responsive, and Easy to Scanwww.nngroup.com · 2021-05-16T16:00:00+0000 — Discusses: How can websites with many categories make it easier for users to choose a relevant entry point?
- 3 Common IA Mistakes (that Are All Due to Low Information Scent)www.nngroup.com · 2023-04-16T16:00:00+0000 — Source used in the article above.
- Information Scent: How Users Decide Where to Go Nextwww.nngroup.com · 2020-02-02T17:00:00+0000 — Discusses: How do cues left by other users affect source selection?; Does a link’s position affect whether users choose it?

