An e-commerce search health check is a ten-point review of your site search that takes about twenty minutes and shows where it is losing sales. It looks at three things: whether shoppers can find and use the search bar, whether search understands what they mean, and whether filters help them narrow a results page down to one product.
Most e-commerce teams put their effort into product pages, ads and checkout. Search gets attention only when someone files a bug. That is a mistake. Shoppers who use site search often have strong purchase intent, and when search fails them they do not file a support ticket. They leave. If you want a hand with any of it, my search and discovery page explains how I work.
The short version: five of the ten checks are critical and three are quick wins (one item is both). Fix the critical ones first. What you find will show where your biggest low-effort revenue gains are sitting.
How to run the check
Open your own site the way a first-time shopper would, on a laptop and then on a phone. Work through the ten items below and mark each one pass, fail or not sure. Type the queries yourself rather than reading a report. The critical items and quick wins are labeled, so you know where to start.
Search bar and basic discoverability
Before search can do anything useful, shoppers have to find it. That sounds obvious, but it is the first place sites slip.
1. The search bar is visible above the fold
Critical, and the most important item on the list. If a shopper has to scroll to find search, most will not bother. On mobile, the icon or bar should sit in the fixed header, not buried in a hamburger menu two taps deep.
2. Search works properly on mobile
Tap targets should be at least 44×44px so the field is easy to hit with a thumb. When the mobile keyboard opens, it should not cover the results or the search field itself. That failure is surprisingly common and makes search unusable on phones.
3. Autocomplete suggestions appear as the shopper types
Autocomplete speeds up the path to a product and steers intent at the same time. A shopper typing “blue” who sees “blue running shoes” and “blue yoga mat” in the dropdown is being guided toward real inventory. Without it, they type a full query and hope for the best.
4. Typo tolerance is active
Critical, and often untested. Type “sungalsses” into your own search bar right now. If you get zero results or a generic error page, you are losing sales every day from ordinary shoppers who cannot spell under pressure. Fuzzy matching is a baseline expectation, not a premium feature. In its 2021 research, Baymard Institute found that 69% of sites don't support autocomplete spelling suggestions for slightly misspelled queries.
Intent and understanding
A search bar that only matches exact strings is a filter, not search. Real search understands what the shopper means, not just what they typed.
5. Synonyms return relevant results
Critical, and frequently broken. If your catalog lists products as “sofa” but a shopper searches “couch,” they should still find them. The same goes for “sneakers” and “trainers,” “TV” and “television,” or any pair of words your customers use interchangeably. Building a synonym library takes an afternoon and pays off immediately.
6. Natural language queries return something useful
Not every site needs full natural language processing, but the more complex your catalog, the more this matters. A search for “gifts for dad under $50” or “waterproof jacket for hiking” should surface relevant products, not an empty page. If your platform supports it, test a handful of intent-based phrases and see what comes back.
7. Zero-results pages offer alternatives
Critical, and a quick win. A dead-end “no results found” page is an exit. A good zero-results page suggests popular products, prompts a broader search or surfaces your top categories. It should never be a wall. Even a simple “here's what's trending” fallback beats nothing.
Filters and navigation
Search gets a shopper to a results set. Filters let them narrow it down to the one product they want. Without good filtering, large catalogs become overwhelming.
8. Filters are available on all results pages
Critical. In my experience this matters once a results set passes about 20 items. Price, size, color, brand and rating are typical, but the right facets depend on your category. A shopper who searches “running shoes,” gets 200 results and has no way to filter by size has hit a dead end.
9. Filters update in real time
A quick win with a clear payoff. When a shopper selects a filter, results should change immediately, with no apply button, no page reload and no waiting. Instant feedback keeps them browsing. Any friction here raises the chance they abandon the page.
10. Active filters are clearly visible
Another quick win. When a shopper has applied filters, say “Size: M, Color: Black, Under $80,” those selections should show as visible tags on the page. Without them, shoppers lose track of what they set, get confused by the results, and sometimes reset everything and start over.
Where to start with your ecommerce search health check
If your site has gaps across all three areas, do not try to fix everything at once. Start with the five critical items: search bar visibility, typo tolerance, synonym handling, a zero-results fallback, and filters on all results pages. These changes cover the most damaging failure modes and are usually achievable without a platform overhaul. The quick wins, real-time filters and visible active filters, can often be a single configuration toggle in your search layer.
Then run through your own search as a first-time shopper and time how long it takes to find a specific product. That number tells you more than any analytics dashboard.
To score your site against all ten points, the interactive e-commerce search audit lets you tick off what your site already does and returns an optimization plan you can share with your team. If you want help acting on what you find, I cover these fixes for e-commerce and marketplace teams.
Common questions about the health check
How long does an e-commerce search health check take?
About twenty minutes. There are ten checks across three areas: the search bar, how well search understands queries, and filters. You run them on your own site the way a first-time shopper would.
Which checks should I fix first?
The five marked critical: a visible search bar, typo tolerance, synonyms, a zero-results fallback and filters on all results pages. They cover the most damaging failure modes. Real-time filters and visible active filters are quick wins.
How do I test typo tolerance?
Type a common word with a slip in it, such as “sungalsses,” into your own search bar. If you get zero results or a generic error page, you are losing the shoppers who cannot spell under pressure.
What should a zero-results page do?
Offer a way forward. Suggest popular products, prompt a broader search or surface your top categories. Even a simple trending fallback is better than a dead-end “no results found” page, which works as an exit.
When do filters become critical?
In my experience, once a results set passes about 20 items. A shopper who gets 200 results for “running shoes” with no way to filter by size has hit a dead end.
Do I need to replace my search platform to fix these?
Usually not. The five critical items are usually achievable without a platform overhaul.
