Blog

How search analytics help merchants: the 5 numbers to check weekly

4 November 2026 · 6 min read

Most merchants check sales, traffic and ad spend every week. Few check their search box. That's odd, because the people typing into it are telling you exactly what they want, in their own words. Search analytics turn those words into a short weekly checklist.

Here are the five numbers we'd look at, what "normal" looks like according to public benchmarks, and what to do when a number looks wrong.

Why your search box deserves 15 minutes a week

Shoppers who search tend to know what they want. Econsultancy reported that up to 30% of visitors use a site's search box, and in the study it cited, searchers converted at 4.63% against a site average of 2.77% (Econsultancy). Prefixbox's 2024 benchmark, based on its own clients' stores, found 16% of shoppers used search but generated 55% of online revenue (Prefixbox 2024 Search Benchmark Report).

Those figures come from different sites and different years, so don't treat them as targets. The point is simpler: a small group of visitors carries a lot of buying intent, and when search fails them, you rarely hear about it.

On Shopify, you'll find most of what you need in Analytics → Reports, in the online store search reports (top searches, searches with no results, searches with no clicks). Report names and availability can vary by plan, so check what your store shows.

Number 1: Search rate

What it is: the share of sessions where someone uses search.

How to work it out: sessions with a search ÷ total sessions. If your reports don't give it directly, a rough estimate is total searches ÷ total sessions. Label it as an estimate, because one visitor can search several times.

What it tells you: a low search rate isn't automatically bad. A ten-product store may not need search. But for stores with large catalogues, or customers who buy by part number, a falling search rate can mean the box is hard to find on mobile, or people tried it once and gave up.

Benchmark: up to 30% per Econsultancy, 16% in Prefixbox's client data (sources above). Your own trend matters more than either.

Number 2: Zero-result rate

What it is: the share of searches that return nothing.

How to work it out: searches with no results ÷ total searches.

What it tells you: this is the most direct measure of search failing. Prefixbox measured an average zero-result rate of 6.04% in the first half of 2024, from 3.66% in DIY to 10.7% in fashion (Prefixbox).

Some zero results are fine. If someone searches for a brand you don't stock, nothing should come back. The ones to worry about are searches for things you do sell. In our own test of 21 Shopify stores, 58 of 62 part-number searches typed without the dash found nothing, even though every product was in stock. We wrote that up in Why Shopify search can't find your SKUs.

And when a search does fail, the page matters. Baymard's benchmark found about 50% of leading ecommerce sites don't give users a useful way forward from a "no results" page (Baymard).

Number 3: Search click-through rate

What it is: the share of searches where the shopper clicks a result.

How to work it out: searches with at least one product click ÷ total searches. Shopify's "searches with no clicks" report shows the other side.

What it tells you: a search that returns results but gets no click is a quieter failure than a zero result. Usually the results are there but wrong. A shopper types fuse box and sees fused spurs. Or a trade word like TRV matches loosely instead of finding the radiator valves.

Benchmark: Prefixbox reported 35.6% of searches led to engagement (a click or add-to-cart) in 2024 H1 (Prefixbox). That's clicks or cart events, so it isn't exactly the same as CTR. Use it as a rough reference.

Number 4: Search conversion rate

What it is: the share of sessions with a search that end in an order.

How to work it out: orders from sessions with a search ÷ sessions with a search. Compare it with your overall conversion rate.

What it tells you: if searchers convert less than everyone else, search is probably getting in their way. The comparison with your own store average is more useful than any industry figure.

Benchmark: Prefixbox reported a 5.4% search conversion rate in 2024 H1, down from 6.3% in 2023 (Prefixbox). Econsultancy's figure was 4.63% (Econsultancy).

Shopify may not split conversion by whether the shopper used search, and what you see can depend on the plan. The online store search reports are about queries. If they don't show orders from sessions that searched, GA4 site-search tracking is a practical workaround. Shopify's results URL carries the query in a q parameter. GA4 can record that as a site search, so you can compare the conversion rate of sessions that searched with your store average. Set that comparison up in GA4. It is not a built-in Shopify report, and it is not one click away in the admin.

Number 5: Top failed queries

What it is: not a rate but a list: the searches that returned nothing or got no clicks, sorted by how often they happened.

What it tells you: this is where the fixes are. The other four numbers tell you whether there's a problem. This list tells you what it is. Prefixbox suggests starting with the top 100 no-result queries and studying the top 10 closely (Prefixbox).

Most failed queries fall into four buckets:

1. Codes typed differently. 221612 for 221-612. See Shopify SKU and part-number search.
2. The shopper's word, not yours. stopcock vs stop tap, TRV vs thermostatic radiator valve. Since Shopify retired synonym groups in Search & Discovery, there's no built-in place to fix these. We covered that in Shopify retired Search & Discovery synonyms.
3. Typos and spellings. Shopify's typo tolerance covers most text fields, but not SKUs (as covered in our SKU post).
4. Things you don't sell. Useful demand data. Maybe you should.

A 15-minute weekly routine

1. Write down the five numbers in a spreadsheet, same day each week. Trends beat snapshots.
2. Open the top 20 no-result searches. Mark each one: code, word, typo, or not stocked.
3. Fix the top three that you do stock. A product tag with the shopper's word or the compact code works on any plan and costs nothing.
4. Re-run those searches yourself, in the dropdown and on the results page.
5. Check next week whether they've dropped off the failed list.

That's it. Three fixes a week adds up to over 150 a year.

What the numbers can't tell you

Search reports count searches. They don't tell you who left angry, or who called your phone line instead. A low zero-result rate can also hide a problem: if a search "finds" the wrong product, it counts as a result. That's why number 3 and number 5 matter as much as number 2.

Also, small stores have small numbers. If you get 40 searches a week, one odd query can swing a rate by several points. Look at the list, not just the percentages.

Where Termatch fits

Termatch does those fixes for you: shopper words and compact codes added as tags, inside your theme's own search, and easy to undo. It is listed on the Shopify App Store. Starter is $9 a month and Pro is $19 a month, each with a 14-day free trial. For a walkthrough of your own failed searches, get in touch.