Skip to content
find sherpasDiscuss your search

Find the search failures
a dashboard can hide.

Six practical probes for your own store. Start with products you know exist, change one thing at a time, and judge what a shopper actually sees.

About five minutes. No setup or analytics access needed.

Choose a known-good starting point.

Pick one important product category and one in-stock product you can verify in the catalogue. Use examples that fit your store and market; the examples below only show the shape of each test.

Compare every variation with the query you already know works. A genuine catalogue gap is not a search failure, and a different result count is not automatically a problem.

Six probes. Change one thing at a time.

Mark what you see after each test. “Usable, but weak” means a shopper can recover; “clear miss” means the experience gives them a credible reason to leave.

1 / 6

Can search find an exact product?

Copy the full name of an in-stock product and search it exactly as written. This is your retrieval baseline.

Known product titleExact-title search
Good signal
The product is first, or immediately visible when genuine variants share the name.
Warning signal
No result, a buried exact match, or unrelated products above it can point to missing searchable data, indexing gaps or an override.

2 / 6

Does one small typo become a dead end?

Take a common product query that works and transpose or replace one letter. Compare the two result sets.

running shoesrunnign shoes
Good signal
The intended category still appears, or one clear correction gets the shopper there.
Warning signal
Zero results or a different product type suggests weak typo tolerance or normalization.

3 / 6

Does search understand customer vocabulary?

Compare your catalogue term with a common synonym or regional expression your customers actually use.

sneakerstrainers
Good signal
Both phrases reach the same intended category without unrelated products taking over.
Warning signal
A large change can expose a vocabulary gap. Check the meaning, not just whether a synonym rule fired.

4 / 6

Is one explicit constraint respected?

Add exactly one visible, verifiable requirement to a working product query—colour, material, audience, size or availability.

hiking bootswaterproof hiking boots
Good signal
The first results satisfy the added requirement and still match the product type.
Warning signal
If the base product matches but the requirement disappears, search is retrieving words without preserving intent.

5 / 6

Does local language survive normalization?

Try one realistic local variation: missing diacritics, plural form, or a word shoppers commonly merge or split.

löparskorloparskor
Good signal
The same product category remains visible with a similarly useful range of results.
Warning signal
A few coincidental matches are not recovery. A thin or off-topic result set still indicates a language-handling gap.

6 / 6

Can a mobile shopper narrow the results?

On a phone-sized screen, run a broad product search and apply one useful filter such as size, price or in-stock status.

Broad result setOne useful filter
Good signal
Filters are easy to find, use plain language and never offer a selectable option that leads nowhere.
Warning signal
A hidden filter control, confusing values or a zero-result combination makes refinement feel unreliable.

0 of 6 probes reviewed

0Looks good
0Usable, but weak
0Clear miss

Complete all six probes to see the pattern.

Judge the visible customer experience, not whether the engine technically returned something.

A useful signal.
Not a complete audit.

This check is designed to expose visible dead ends quickly. It cannot tell you how often they happen, which catalogue segments are affected, or whether a fix improves the rest of your query set.

A full audit repeats these probes systematically across real query patterns, products, languages and customer intents. It separates retrieval, interpretation, ranking and interface failures before prioritising the fixes by customer impact.

What should your
search do better?

Tell us about your store, your search platform and what you want to improve. We will work out where it makes sense to start.

Discuss your search