Online store search: why "no results" costs you more than you think

In most online stores, the search box is the most neglected part of the buying journey. The paradox is that it's used by the customers with the highest purchase intent - the ones who know exactly what they want. And it's precisely them the store most often answers with: "no results".

A customer who can't find it won't ask - they'll just leave

When a customer in a physical shop can't find something, they ask a sales assistant. In an online store that conversation usually doesn't happen: the user types a phrase, sees an empty results page and goes back to Google - most often straight to a competitor. They leave no complaint or email behind. Only a trace in the analytics that almost nobody looks at: a session ending on the search results page.

That's why the problem is so underestimated. The cost of "no results" doesn't show up as a line in the profit and loss account - it's scattered across hundreds of silent exits from the store every month.

Where "no results" comes from

A standard search engine matches words, not intent. Any one of the following is enough for a customer to see an empty page:

• Typos and word forms. "Monobloc heat pmup" instead of "monobloc heat pump" - for classic matching, that's a completely different phrase.
• Colloquial names. The customer searches for "a Junkers", the catalogue only knows "gas water heater".
• Symptoms instead of products. At Ciepło24.pl, customers typed phrases like "boiler not heating" - no product page contains those words.
• Languages and units. A phrase in another language, dimensions in inches, manufacturer codes with or without a hyphen - each variant can "lose" the product.

The more technical the range, the worse it gets: the number of name variants grows, and the chance that the customer hits the exact wording in the catalogue falls.

What it really costs

A simple calculation on data you'll find in your own analytics. Check three numbers: how many sessions a month use search, what percentage of queries end with no results, and what the conversion rate from search is compared with the store as a whole. Before implementation, the typical picture at our clients was: somewhere between 10 and 20% of queries with no results, and search sessions abandoned almost immediately after the empty page.

Then there's the cost on the team's side. Some customers, instead of leaving, call or write: "do you have…?". The Ciepło24 customer service team was answering the same questions dozens of times a day - questions the search results page should have been answering.

What semantic search changes

Semantic search (in our case: Elasticsearch plus an AI layer in ecSearch) doesn't compare letters, it compares meaning. In practice that means four things:

• Typos stop being a problem - "monobloc heat pmup" returns exactly what it should.
• Intent instead of phrase - a descriptive query leads to the right category, even when the words don't appear on the product page.
• Suggestions after 3 characters - the customer picks from relevant suggestions before they have a chance to make a mistake.
• Auto-translation - the customer searches in their language and finds it in your catalogue. For one of our UK clients (Wholesale Cakes), basket value rose by 38%.

Results at Ciepło24.pl after implementing ecSearch: +45% CTR on search results and +25% conversion from search. Implementation - with a standard catalogue - takes 2 weeks, because it doesn't require changing the store's engine.

How to check your store in 15 minutes

1. Pull a zero-results report from your analytics or search panel. If your search engine doesn't have one - that's already an answer.
2. Run a typo test: type your 10 best-selling products with a mistake in the name. Count how many times the store showed an empty page.
3. Compare the conversion rate of sessions with and without search. If sessions with search don't convert noticeably better, your search isn't doing its job.

These three steps show the scale of the problem - and they're exactly where we start every search consultation.

FAQ