Product discovery and a delivery question need different answers.
Search types: names, SKUs, attributes and help
A query may name a product, a feature or a question about the store. Provide a relevant destination for each.
Research context: Baymard · E-commerce search query types ↗
Start with the query types your assortment needs. Avoid promising arbitrary intent support.
Try the pattern
Task and expected result
- 01
Search NORD-02: only Nord Tur should appear. Search termokopp: both mugs should appear.
- 02
Search blue 500: Nord Tur should match one variant. Search green 500: combining two variants must not produce a match.
- 03
Search returns: show the help answer, not an unrelated item. Search unknownitem: show zero results and specific recovery paths.
How to investigate value
Observe whether participants find an item meeting every need, not just how many cards they click. For measurement: relevant selections per eligible search/list visitor.
Watch for side effects
Zero results, false variant matches, lost list context and slow filters. More product-page visits are not necessarily better.
This is a proposed investigation, not a measured result. Choose one hypothesis and define the audience before testing.
Build a measurement plan for this concept ↗