Aglet

Investigate result sort expectations with ranked comparisons

An investigation should show which input determines order and where the visible sort promise diverges. Replay search and category contexts, change one sort or rule at a time, and compare representative positions. Keep result identity stable so a changing catalog does not obscure the comparison.

Build a useful investigation brief

  1. Build the order table

    For each query, record route, sort, default, filters, top results, rank factors, and visible labels. Include relevance, date or price sorts where available, plus a category path. Note any promotion or pin that affects the order.

  2. Isolate one input

    Hold query and scope fixed while changing sort, then hold sort fixed while changing category or rule. Compare displaced results and label state. Record a deliberate exception as evidence and a silent mismatch as a separate finding.

  3. Write the sort brief

    State the first input that changes order, affected context, and competing explanations. Recommend one bounded experiment with a pass condition based on user expectation. Do not infer relevance from position without the selected sort and result policy.

What to carry forward

The investigation is complete when order changes can be reproduced from inputs and the visible promise has a named boundary. Stop with a single-variable experiment or a clear policy explanation. Keep category and global search separate when they use different ordering rules.

Technical background: Elastic documentation.

Keep the decision with the work.

Use a Work Item in Aglet to record the problem, the evidence you have, and the next decision. Add an owner and priority, then keep updates in the discussion so the next person can follow the reasoning.

Create an account See the product workflow