Your most valuable report is the list of searches that found nothing

Every search bar records what gets typed into it. Most organisations look at their most popular searches — which feels sensible, but it is the least useful list you can make. What gets found often is, by definition, already fine.

The interesting list is the other one: the searches that returned zero results. Those tell you two things at once, and it pays to keep them apart.

Missing content

If ten colleagues search for "facade evening light" and nothing comes back, that is a briefing. Not a feeling that something might be missing, but a counted request from ten people who needed it and did not find it.

Badly described content

More often the imagery exists but goes by another name. The photographer called it building_02.jpg, the AI layer saw "architecture", and nobody searches for architecture. That is not a content gap but a language problem — solved with a synonym, not with a new shoot.

You tell the difference by checking whether anything nearby exists. That is why sökbar shows semantic matches even on an empty result: if something turns up there, you know it is a language problem.

Why almost nobody looks

Because it is a report that proves you wrong. A list of things your people could not find reads like a list of mistakes. But it is the cheapest research you will ever run: your users have already filled it in for you.