The day’s workbench tool is information scent, a useful idea from Information Foraging Theory. Peter Pirolli and Stuart Card argued that people seeking information behave, in part, like organisms searching for food. They follow cues that suggest whether a richer patch lies nearby. A headline, link label, source name, abstract, snippet, or table of contents can all act as scent.

The idea is practical because it gives knowledge workers permission to quit weak trails. If a source promises relevance but keeps failing to produce evidence, the rational move is not always to dig harder. Often it is to leave the patch, sample another trail, and return only if stronger cues appear. Research quality improves when attention is treated as a scarce resource rather than a moral test of perseverance.

For writers, information scent is a design obligation. A reader should quickly know what a document contains, why it matters, and where the useful parts live. Vague section titles, buried conclusions, decorative introductions, and generic links all destroy scent. The reader may not be lazy; the trail may simply be badly marked.

For operators, the same rule applies to meetings and project work. “Q3 review” has weak scent. “Q3 review: decide whether to pause enterprise launch” has strong scent. It tells participants what decision is at stake, what evidence to bring, and whether attendance is worth the interruption. The best agendas are not polite labels. They are navigational instruments.

For AI-assisted work, information scent becomes even more important. Agents can search broadly and produce fluent summaries, but they can also follow low-quality trails at scale. A good prompt, source list, or task brief should include cues about authority, recency, acceptable evidence, and exit criteria. Without those cues, the system may spend inference budget polishing a weak patch.

The habit to file is simple. Before going deep, scan for scent: source quality, specificity, date, named evidence, and whether the next click is likely to answer the question. While writing, add scent for others: concrete titles, honest summaries, useful links, and ledes that say what changed.

Attention is not a warehouse. It is a field operation. The skilled worker samples broadly, follows strong traces, and leaves the empty ground without regret.