Every search a shopper types into your store is a short, unfiltered sentence about what they want. Most of those searches end well. A smaller number return nothing, and those are the ones worth your time: each one is a person who came to buy, told you exactly what they were after, and left with an empty page.
This guide shows how to collect that data, how to find the zero-result searches in it, what to do about each kind, and how to turn the whole thing into a short weekly routine. It is about reading your own store’s search data, not about choosing a search plugin. If you are still choosing one, our list of WooCommerce search plugins covers that side.
Why search data is worth a weekly look
Browsing and searching are different behaviors. A shopper who browses follows the paths you built: categories, menus, banners. A shopper who searches skips all of it and states a need in their own words. That makes the search log the closest thing a store has to a free, continuous customer interview.
The log answers questions that sales reports cannot:
- Which words do customers use for products you already sell, and do those words match your product titles?
- Which products do customers expect to find that you do not stock?
- Which spellings, abbreviations and brand names trip up your current search?
- Which searches lead to a result page that people immediately abandon?
Sales reports only describe people who found something to buy. Search data describes the ones who did not. Both matter, but the second group is invisible without a deliberate effort to look.
Where search terms come from
There are three practical sources. You can use one or all of them. Each has blind spots, so it helps to know what each one actually records.
The WooCommerce search query (s)
When a visitor submits the product search form, the browser requests a URL with the search text in the s query parameter, usually alongside post_type=product. Your server access logs and any analytics tool that reads URLs can therefore see the term. This is the foundation of everything else in this guide: if a search happened, the term was in the URL.
On its own, though, WordPress does not keep a history of searches. Nothing in a stock WooCommerce install saves the terms for you to review later. You need either an analytics tool that captures the URL or a search plugin that stores its own log.
GA4 site search
Google Analytics 4 can record site searches through enhanced measurement. According to Google’s documentation for enhanced measurement events, the view_search_results event fires when a user is shown a search results page, as indicated by a URL query parameter. By default it looks for one of five parameters: q, s, search, query and keyword. WooCommerce uses s, so it is covered without configuration.
The term itself is stored in the search_term event parameter, which populates the Search term dimension in GA4. To check that it is working:
- In GA4, open Admin, then Data streams, and select your web stream.
- Open the enhanced measurement settings and confirm that Site search is switched on.
- Run a search on your store, then look in the Realtime report for a
view_search_resultsevent. - After a day, build a table in Explore with Search term as the row and Event count as the value.
What GA4 does not tell you is whether a search returned any products. It records that a results page was shown, and an empty results page is still a results page. For the zero-result question you need one of the next two options.
A search plugin’s own report
Some search plugins keep a log of queries and mark the ones that found nothing. This is the most direct route to zero-result data because the plugin knows how many results it returned.
FiboSearch is one verified example. Its announcement of FiboSearch Analytics describes an Analytics tab in the plugin settings, available in both the free and Pro versions from version 1.19.0, aimed at showing what customers type and which phrases return no results. The vendor also says the tool analyzes search phrases only and, unlike Google Analytics, does not store customer or order data. We have not checked other search plugins for the same feature, so if you use a different one, look through its settings for a statistics or analytics tab before adding anything else.
If your plugin has no report, you can still capture the zero-result case yourself, as shown below.
Capturing zero-result searches without a plugin report
WooCommerce fires the woocommerce_no_products_found action in its product archive template when a query returns nothing. A small snippet can hook it, and when the page is a search, send a custom GA4 event:
add_action( 'woocommerce_no_products_found', function () {
if ( ! is_search() ) {
return;
}
$term = wp_json_encode( get_search_query() );
echo '<script>window.gtag && gtag("event", "search_no_results", { search_term: ' . $term . ' });</script>';
} );
This assumes GA4 is already loaded through gtag.js. If you load it through Google Tag Manager, push an equivalent event to the data layer instead. After a day of traffic, the search_no_results event gives you a ranked list of failed searches in the same GA4 property as everything else. Test it on a staging copy first, and check that your consent banner allows the script to run.
Finding the zero-result searches
Once data is flowing, build the list you will work from. Export two things for the last 30 days:
- All search terms with their counts.
- All zero-result terms with their counts.
Sort both by count, highest first. Then add a note column. A short spreadsheet is enough. The point is to review the top 30 or 40 rows, not to analyze thousands.
Two cautions before you read the list:
- Ignore one-off noise. A term searched once, with a typo and no obvious product, rarely deserves action. Look at the repeated terms first.
- Merge near-duplicates. “hoodie”, “hoodies” and “hooded sweatshirt” may all be one demand. Add their counts together before deciding how important it is.
Why searches fail, and the fix for each cause
Go through the zero-result list and label each row with a cause. Nearly every row falls into one of the groups below. The label tells you the fix.
| Cause | What you see in the log | Fix |
|---|---|---|
| Different word for the same product | “sneakers” returns nothing, you sell “trainers” | Synonyms, or add the word to the title or short description |
| Plural, spelling or typo | “sunglases”, “cabels” | Fuzzy matching in the search plugin, or a redirect for repeated typos |
| Attribute searched as a keyword | “red”, “xl”, “cotton” | Make sure attribute values are searched, or send these to filtered pages |
| Category or brand name | “kitchen”, a brand you carry | Redirect to the category or brand page |
| SKU or model number | “AB-1234” | Confirm SKUs are searchable, or redirect to the product |
| Product you do not stock | A specific item with real, repeated demand | Treat as a sourcing signal |
| Content, not product | “returns”, “shipping”, “size guide” | Redirect to the page that answers it |
Fix 1: Synonyms
Synonyms are the highest-value fix for most stores because they solve the “different word” problem without changing the catalog. If your search plugin has a synonym list, add the customer’s word and point it at yours. Keep the list short and driven by the log, not by guesses. A synonym list that grows without review starts producing odd results.
If your search has no synonym feature, you can still work with the defaults. The default WordPress search looks at titles, excerpts and content, so you can put the customer’s word where it will be found. A short phrase in the product’s short description, such as “Also known as trainers”, is natural enough to read and indexed by default. Do not stuff a hidden block of keywords into the page. It looks bad to shoppers and does little for search.
Fix 2: Product titles
Titles are what shoppers read on the results page, and the default WordPress search covers them. If many people search for a term and the right product exists but ranks poorly or not at all, the title is often too internal. “Model 4 Series” says less than “Stainless Steel Water Bottle, 750 ml”. Rewrite titles around the words customers use, not the words in your supplier’s spreadsheet.
Do this carefully for products that already rank in Google, since a title change can shift the page title too. Test with a handful of products before editing a large batch.
Fix 3: Tags and attributes
Product tags and attributes are your structured vocabulary. They help in two ways. First, they power filtered pages, so a search for “red” can be answered by a color-filtered archive. Second, many search plugins can index tags, categories and attributes alongside product text. Check your plugin’s settings for which fields it searches, because defaults differ.
When the log shows people searching for an attribute value, check three things:
- Does the attribute exist on the products that have it?
- Is the attribute used consistently (“XL” and “Extra Large” are not the same value to a database)?
- Does the search index include attributes?
Fixing inconsistent attribute values is dull work, and it pays off in both search and filtering.
Fix 4: Redirects
Some searches are really navigation. Someone typing a category name, a brand or “shipping” wants a specific page, and a list of loosely matched products is a worse answer than the page itself. A redirect for the top few of these is cheap and effective.
If your search plugin supports keyword redirects, use that. If not, a minimal snippet works for a short list:
add_action( 'template_redirect', function () {
if ( ! is_search() ) {
return;
}
$map = array(
'shipping' => '/shipping-and-delivery/',
'returns' => '/returns/',
);
$term = strtolower( trim( get_search_query() ) );
if ( isset( $map[ $term ] ) ) {
wp_safe_redirect( home_url( $map[ $term ] ) );
exit;
}
} );
Keep the map small. If it grows past a dozen entries, move it into your search plugin or a redirect manager so a non-developer can maintain it. Also remember that a redirect hides the failed search from the log, so keep a note of which terms you redirected and why.
Fix 5: Typos
Do not fix typos one by one unless a specific misspelling shows up repeatedly. A search engine with fuzzy matching handles most of them. Where one misspelling dominates, usually because it is the common way a brand is spelled, add it as a synonym.
Spotting demand for products you do not stock
After you have cleared the vocabulary problems, the remaining rows are the interesting ones. These are specific items, searched repeatedly, that you do not sell.
Not every one is an opportunity. Use a simple filter before you act:
- Repeat demand. The term appears across several weeks, not in one spike.
- Fit. It sits next to what you already sell and serves the same customer.
- Spread. Many different visitors searched it, not one person trying five times.
- Supply. You can source it at a margin that works.
A term that passes all four deserves a small test. Add one product, or a pre-order listing if that suits your business, and watch whether searches turn into orders. A term that fails the filter goes on a watch list. Review the watch list again next month, and drop anything that has not come back.
There is also a defensive use. If many people search for a product you discontinued, a short page explaining the replacement is better than a dead end. A redirect from the old term to the successor product works here too.
Reading search behavior beyond the zero-result list
Zero-result terms are the clearest signal, but they are not the only one. Two other patterns are worth checking when you have time.
Searches with results that nobody clicks. If your analytics shows the term but the visitor leaves from the results page, the results were probably technically correct and practically wrong. Look at the first three results for that term. Are the right products at the top? Is the image or title unclear? This is a ranking and presentation problem, and a different one from zero results.
Repeated searches in one session. A visitor who searches “wireless charger”, then “charger pad”, then “qi charger” is telling you your first results did not match. Those sequences point to a synonym gap or to a catalog that uses unfamiliar words.
A quick comparison of how each type shows up in your tools:
| Signal | Where to see it | What it usually means |
|---|---|---|
| Zero-result term | Search plugin report, or the custom search_no_results event |
Vocabulary gap or missing product |
| High count, low click-through | GA4 events combined with page exits | Weak ranking or unclear results |
| Same visitor, several rewrites in a row | GA4 sequences of view_search_results |
Synonym gap |
| New term appearing suddenly | Week-over-week comparison | Trend, season or a campaign you ran |
A weekly review routine
A routine beats an occasional deep dive. Search data has a short half-life, and small weekly fixes add up. Set aside thirty minutes at a fixed time each week and follow the same steps.
- Export the last 7 days of zero-result terms, plus the top 20 overall search terms.
- Label each row with a cause from the table above.
- Make the quick fixes first: add synonyms, correct titles, add redirects. These take minutes.
- Move missing-product rows to a separate sheet and add this week’s counts to the running totals.
- Re-run last week’s top failed terms on the live store to confirm your fixes worked.
- Write one line at the top of the sheet: what changed and which term to watch next week.
Step 5 is the one most people skip, and it is the one that proves the work. If a term that returned nothing last week now returns the right product, the fix worked. If it still returns nothing, the cause label was wrong and you should try the next likely fix.
Once a month, spend a little longer: look at the watch list for sourcing candidates, trim synonyms that no longer match anything, and check whether the share of zero-result searches in your total searches is going down. Do not chase a target of zero. Some failed searches are unfixable, such as random strings, and that is fine.
Common mistakes
- Treating the log as a to-do list. Not every failed search deserves a fix. Rank by repeated demand and act on the top of the list.
- Over-building synonyms. Each synonym can pull unrelated products into results. Add them when the log shows a need.
- Changing titles in bulk. Titles affect the product page’s appearance in Google and in feeds. Make changes in small batches and watch the effect.
- Redirecting everything. A redirect is right for navigation-style terms. For product searches, a good results page is better than a bounce to a category.
- Forgetting consent. If you send search terms to analytics, check that your privacy policy and consent banner cover it. Shoppers sometimes type personal details into search boxes.
- Never confirming the fix. Re-test last week’s failures. Without that, you cannot tell which fixes work.
Making search data part of how you run the store
The same list helps outside search. The top terms can guide what goes on the homepage. The content-style terms show which questions the site does not answer, and the missing-product list is useful when planning orders. If you already use a broader reporting tool, our guide to WooCommerce analytics plugins shows how other reports sit alongside search data.
Presentation matters too. A well-tuned search behind a cramped header box is easy to miss, so make the box obvious and test the results page on a phone. Reign is a WooCommerce-compatible theme if you are reviewing your header layout.
Frequently asked questions
Does WooCommerce record what customers search for by default?
No. A stock WooCommerce install does not save a history of search terms. The term is present in the URL, in the s parameter, so analytics tools that read URLs can capture it, and some search plugins keep their own log. If you want a record, set up GA4 site search or enable your search plugin’s report.
How do I see search terms in GA4?
Turn on Site search under enhanced measurement for your web data stream. GA4 then sends a view_search_results event whenever the URL contains one of the default query parameters, which include s, and stores the term in the search_term parameter. Add Search term as a row dimension in an Explore report to read the list.
Can GA4 tell me which searches returned no results?
Not by itself. GA4 records that a search results page was displayed, whether or not products were on it. To capture zero-result searches, use a search plugin that reports them, or send your own event from the woocommerce_no_products_found action as shown above.
How many zero-result searches is too many?
There is no universal benchmark worth quoting. Track your own share of zero-result searches over time and look at the repeated terms. A downward trend after you add synonyms and fix titles is the result to look for.
Should I redirect zero-result searches to a category page?
Redirect when the term clearly names a category, a brand or an information page, because that is what the shopper wanted. For product searches, a redirect can hide the real problem. Fix the vocabulary or the title so the right product appears in results.
How often should I review search data?
Weekly for the zero-result list, using a fixed thirty-minute slot, and monthly for the longer-term items such as sourcing candidates and unused synonyms. Smaller stores with little search traffic can review every two weeks, since there is less new data each week.