A nav label must accurately preview what it opens

What it is

The word on a navigation item has to let people judge what sits behind it before they click. Accurate preview means the label and the destination set match: “Pricing” opens prices, not company news that happens to mention a price. Information scent on navigation is that preview. The word can be short; it cannot send people into another kind of object. Failed preview is not unattractive copy. It is a mismatch between the set the label names and the set arrival delivers.

This is “which kind of content this slot leads to.” It is not whether two slot names are easy to tell apart, nor whether the site uses one word everywhere.

Why it happens

The click happens before the open. People treat the label as a sample and infer the centre of a hidden set. If the sample is from the set’s edge (“Resources” mixing jobs, blog posts, and PDFs), the inferred centre is biased and the first arrival is already a falsification. After falsification, scent on similar labels is downweighted; the next decision of that kind trusts the label less and opens to look.

Accuracy is not copying the directory onto the label. It is: most of the objects a person would think of under this word are in the destination, and the main objects in the destination are coverable by this word. Too narrow or too wide is still a failed preview.

Studying it

Have people predict content from the label alone, plus first clicks; do not ask whether the label “sounds nice.”

• Paradigms: reverse cards—show only the nav label, ask what will be behind it and for three examples; compare with the real destination set. Tree tests or first clicks for whether tasks that enter this label found the right class.
• Independent variables: specificity of the label, purity of the destination set, whether a department name is used as the preview.
• Dependent variables: overlap of predicted and real sets, first-click accuracy, immediate back-outs after open.
• Methodological note: completing the task after opening does not prove preview accuracy—people may have probed until they hit. Score “surprised on arrival” separately. Staff know backend module names and inflate accuracy.

Where it stops holding

Deliberate overview destinations (“Explore,” “Discover”) preview a mix; the accuracy bar becomes “this is mixed browse,” not one object type. If people use them as exact entrances, overviews should not occupy a precise slot on primary nav. Search suggestions and nav labels are different previews: suggestions may follow the query; a nav label keeps the same preview on every page. A legal document’s official title may appear after arrival; the entrance label should still be the lookup preview.

Applying it

• Write one sentence per nav item: “the bulk of what a click opens is ___.” Check it against a real object list. Mismatch: change the label or the membership.
• Do not let words that almost any object can enter (“Hub,” “Resources,” “Solutions”) occupy slots that need a precise preview.
• Verify by covering the page, leaving only nav labels, and asking someone uninvolved for two examples they would see. Examples that fall outside the real destination, or inability to name the real bulk objects, means the preview is inaccurate.

Related

• Within the group: G2.08.2 A mismatch between expectation and arrival lowers trust in the whole site · G2.08.3 Vague labels force people to probe items one by one
• Adjacent: G1.04 Labeling systems · T2.02 Predictability of titles and labels · G2.01 Hierarchical navigation
• Search terms: information scent · navigation label · scent accuracy