Internal search is one of the most neglected components in most online stores. It usually ships native with the platform, gets assumed to be “good enough,” and nobody looks at it again. The measurable reality: visitors who use the search bar convert 2 to 5 times more than those who browse by category, and a bad search bar leaves that gap on the table every single day.
What the data says about internal search
Consistent studies over the last few years (Baymard, Forrester, various search SaaS vendors) report three relevant data points:
- Between 15 and 30% of a store’s visitors use the internal search bar at least once per session.
- Visitors who use search convert at a rate 2-5 times higher than those who only browse.
- 40-60% of searches that return “no results” correspond to products the store actually sells, just under a different synonym or with a typo.
That last point is the most painful one. Every “no results” search for a product available under another name is a lost sale from a customer who already showed clear intent.
Signs of a broken search bar
Four indicators worth checking before assuming your search works fine:
- No-results search rate. If it’s above 10-15% of total searches, there’s a problem. A good search bar returns something useful even for imperfect queries.
- Exit rate after searching. If more than 40% of sessions with a search end without viewing a product page, search isn’t converting intent into clicks.
- Searches that don’t find what the customer wanted. Identifiable by cross-referencing search queries with support queries (“I searched for X and it didn’t show up”).
- Time from search to product click. If it’s over 15 seconds, the results aren’t relevant and the customer is digging.
If two of these four symptoms show up, your search bar is a candidate for intervention.
What a genuinely good search bar does
Five capabilities that separate a functional search bar from one that sells:
- Live autocomplete. Instant suggestions with products and categories as the user types. Cuts search time and shows the customer you have what they’re looking for before they hit enter.
- Typo tolerance. “sneakrs” should return the same as “sneakers.” Without this, you lose every fast-fingered search.
- Synonyms and related terms. “laptop,” “notebook,” “portable computer” all mean the same thing. “hoodie,” “hooded sweatshirt” too. Your customer’s language isn’t always your catalog’s language.
- Attribute search, not just name search. If you sell clothing with sizes and colors, a search like “black trousers size L” should filter by name, color and size at once.
- Redirects for specific searches. For strategic queries (“black friday,” “outlet,” “new arrivals”), redirect straight to the relevant page or category instead of listing products.
Almost no native search bar on a standard platform implements all five. Some cover autocomplete and attributes; synonyms and typo tolerance are usually extensions.
Native search vs external SaaS vs module
Three options for improving internal search:
- Improve the native search. With good aliasing, indexing and filters, native search can be enough for small stores. Past roughly 1,000 products it tends to fall short.
- External SaaS (an external SaaS-type search engine). Replaces your search with an external service. Excellent result quality, usage-based pricing (typically starting around €30-100/month depending on volume). Implementation takes half a day to several days.
- A dedicated module on the platform itself. Improves native search with autocomplete, synonyms and typo tolerance while keeping everything inside the store. No recurring cost, more limited than SaaS but enough for most catalogs.
SaaS wins on pure quality; the module wins on cost over time and operational simplicity. For catalogs up to 10,000 products with average traffic, the dedicated module usually has the better cost/benefit ratio.
Zeyvro has a module of this kind for PrestaShop 8: the Smart Search product page has the full detail of what it includes.
No-results searches: an opportunity, not a failure
When a customer searches for something and doesn’t find it, there are three reasonable actions:
- Show related categories to keep exploring. “We couldn’t find X, but here are the categories where you’re most likely to find something similar” — you don’t need to guess the exact product, just avoid leaving the customer on an empty page.
- Offer to save the search. An email to notify the customer if the product enters the catalog. Useful for discontinued products that might come back.
- Log the query. If the same product is searched for repeatedly and not found, that’s a demand signal. Data for catalog decisions.
What doesn’t work: returning an empty page with “No results found” and a button back to the homepage. That’s assisted abandonment.
Next step
The next post in this series is the tutorial for setting up internal search with autocomplete, synonyms and redirects in PrestaShop 8. If you’d like a review of your current search before investing in changing it, write to us at hola@zeyvro.com with the URL.