Why your platform's native search engine isn't enough

Most e-commerce sites set up their search engine once, at launch, and never touch it again. That's understandable: internal search doesn't get watched the way homepage bounce rate or checkout flow does. Yet it's often where a meaningful share of lost sales happens.

Here are the three blind spots that come up most often when auditing a native engine.

1. Typo tolerance exists, within limits

The three platforms differ here, and none of them ships with nothing. According to Shopify's documentation, its search accepts one wrong letter or two swapped letters, only if the first four letters are right, and only in the title, product type, variant title and vendor. PrestaShop turns approximate search on by default: PS_SEARCH_FUZZY is 1 in the shipped configuration of versions 8.2.8 and 9.1.5, with a minimum word length of 3 (PS_SEARCH_MINWORDLEN). WooCommerce, in its source code (version 11.1.0), matches each word as a substring of the title, excerpt and content, and that query has no typo tolerance.

So the gap isn't the absence of tolerance, it's where it stops. "swetaer" for "sweater" swaps letters inside the first four, which Shopify requires to be right. "M8" is shorter than PrestaShop's three characters. On a technical catalog, short terms and codes are exactly what falls outside these limits.

2. A sort order that ignores stock and margin

By default, most engines sort results by raw text relevance, or worse, by catalog add date. Result: an out-of-stock product can show up first, while an available, more profitable item ends up on page three.

A sort order that factors in stock, margin, and actual relevance directly changes the conversion rate of search results pages — without touching the rest of the site.

3. Missing synonyms and business vocabulary

Your customers don't always speak your catalog's language. A shopper might search "jumper" where your product sheet says "sweater," or "trainers" where you wrote "sneakers." Without a synonym dictionary tailored to your business, each of these searches returns zero results — even though the product genuinely exists.

A search engine that returns "no results" for a product you have in stock isn't a technical detail — it's a lost sale, every single time it happens.

How to check where you stand

Three tests are enough to get a first read, directly on your own site:

• Search for a product you know is in stock, with a deliberate typo
• Search a generic term from your sector ("sweater," "shoes"...) and look at what comes up first
• Search a common synonym of one of your flagship products rather than its exact name

If any of these three tests fails, that's a concrete signal — not a matter of opinion about design or UX, but a measurable problem costing you sales every day.

Two more signals are read in your statistics rather than at the keyboard: the share of searches with no result, and the conversion of visitors who search. They are among the five signs your search engine is costing you sales.

Corrected on September 15, 2026. The first section said native engines have no typo tolerance and that it is almost never enabled by default. That was false for Shopify and PrestaShop, which both ship with it turned on.