Consistent product names do not solve every search mismatch. A customer may use a valid term that differs from the catalog. Investigate real queries, then configure available search features. This guide covers synonyms and normalization beyond the basic naming pattern.
Build a test set from real questions
Collect zero-result queries and service questions without names or phone numbers. Classify them as familiar terms, spelling variants, product codes or requests for items the store does not sell.
Not every zero-result query indicates a search bug. A shopper asking for an absent category needs a clear explanation or relevant alternative, not a distant match that implies the product exists.
Test normalization without damaging displayed names
Test Arabic letter variants, diacritics, spacing and numeral forms that occur in your audience’s queries. A correct display title can coexist with a normalized search representation if the engine supports it.
Do not apply broad replacements to model codes indiscriminately. A character or hyphen may distinguish two products. Preserve the original identifier and an exact-match path when needed.
Search test matrix
Scroll the table horizontally to see all columns.
| Case | Example | Decision |
|---|---|---|
| Familiar term | Two local words for a bag | Review a category-specific synonym |
| Different specification | Leather / synthetic leather | Do not treat them as one fact |
| Identifier | ABC-10 / ABC-100 | Preserve the distinction |
Treat synonyms as deliberate relationships
“Backpack” and a familiar local term may refer to the same category. Leather and synthetic leather are not interchangeable specifications. Have a category specialist review relationships and decide whether a one-way or two-way expansion is appropriate.
Use synonym configuration if available, or improve descriptions and categories naturally. This guide does not claim that every Mollkom account has a synonym console or automatic spelling correction; verify the actual search capabilities.
Measure false matches as well as missing results
Rerun the query set after each small batch and record expected items and irrelevant matches. Fixing one term should not break queries that already worked. Keep changes small enough to reverse.
Review product selection after search and query reformulation while respecting privacy and sample size. More clicks on the wrong items followed by immediate returns can signal poorer relevance rather than improvement.
Common questions
Should I create a title for every spelling?
No. Keep one clear display name and address query variants through suitable fields or search configuration.
How do I spot a bad synonym?
It introduces irrelevant items or breaks queries that previously worked. Evaluate a query set rather than a single example.
Your next step
Continue with the product naming guide, or register and start your store. Apply the process to a small sample before expanding.
Sources and references
Information reviewed on 27 September 2026. Service availability and provider terms can change.