Let’s say you run a clothing store. A customer, Jonah, comes in to buy a t-shirt. In the search bar, he types in: “V-neck t-shirt.” He gets 726 results. That’s a lot more than he can sift through, so he refines his search term. “Instead of “V-neck t-shirt,” he searches for “Black V-neck t-shirt.” The results are narrowed down to 87. If Jonah wants to narrow down the search further, he can type in a third prompt: “Black v-neck cotton t-shirt.”

Jonah’s experience is not unique. The average customer uses 2.2 rounds of search iteration to get what he wants. They type in the first term and if the results are too broad, type in additional words to get them the desired results. Customers do this two or three times, sometimes more.

But as you can see with Jonah’s example, customers don’t change their search terms entirely. They usually just add to them. Jonah started with “v-neck t-shirt,” then added “black” to that term to get “Black v-neck t-shirt,” and finally added “cotton” to get “Black v-neck cotton t-shirt.”

Like with most customers, Jonah just adds new words to the existing query. It’s just a lot easier for him. If Jonah had to retype the whole query all over again whenever he wanted to refine the search, he might start getting annoyed at the unnecessary effort.

That is why you persist the search query on the results page. It allows customers to easily refine their search terms. They can add new words, remove unnecessary ones, and correct misspelled ones if they don’t get the desired result. Having to retype the entire search query every time they want to tweak their search results is just taxing.

Two thirds of e-commerce sites tested by Baymard persist search queries in the results page. If yours is one of the one third that don’t, fix it. A smooth search experience makes for happy customers. And I’m yet to meet anyone who hates a happy customer.