Zero results for termokop can suggest that the store sells no mugs.
Typos: suggest corrections and keep the query
Offer a recognizable correction for known typos. Retain the original query until the shopper chooses it.
Research context: Baymard · Misspellings in autocomplete ↗
Small catalogs can use a controlled dictionary. Avoid guessing corrections for SKUs and model numbers.
Try the pattern
Task and expected result
- 01
Search termokop: see an optional termokopp suggestion.
- 02
Choose the correction: update the field and URL.
- 03
Search an unknown SKU: never substitute a different product number.
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 ↗