Contents 4
Sufficient consensus
Interactive information retrieval studies search as a sequence of queries, results and reformulations rather than a single lookup. Session analysis, exploratory search and query reformulation are its core topics.
Real search behaviour is iterative and it is well documented.
If long research tasks were ordinary sessions, their share of queries would match their share of sessions. It is more than two and a half times as large.
| Sessions spent on a research mission | Share |
|---|---|
| Of all search sessions | 10 % |
| Of all queries | more than 25 % |
Donato, D., Bonchi, F., Chi, T. and Maarek, Y., Do you want to take notes? Identifying research missions in Yahoo! Search Pad, WWW 2010. Editors reviewed more than 7,000 user sessions, each covering 3 days of activity.
What the result pages leave out
These belong to the subject. People ask about them. They are missing from the agreed coverage.
- Content for the second query. Sites are built for the entry query. The reformulation that follows it usually lands on somebody else's page.
- Learning during the search. The searcher at query three knows more than the searcher at query one and needs a different text.
- The dead end. Sessions that end without an answer are the most informative ones.
- Sessions across days. A decision can take three weeks of occasional searching. Session metrics cover minutes.
- Handing over. A page that tells a reader what to search next keeps the relationship even when it cannot answer.
What people actually want to know
- What will this reader type next?
- What did they type before they got here?
- Where should I send someone whose question I cannot answer?
- How long does this decision really take?
A site that answers the next question, including the one that leads away, is the one people come back to.
Last revised 2026-09-16.