Your Search Can Be a Barrier to Product Discovery

Search is supposed to help shoppers get to the right product faster. For one e-commerce client, that was not always happening.

The site had a prominent search experience visually - but when we reviewed the experience through an accessibility lens, we found that some of the most important parts of search were difficult to navigate without a mouse. 

The Challenge

The search experience for the client’s platform depended heavily on visual interaction.

As users typed, autocomplete suggestions appeared dynamically beneath the search field, along with product recommendations, categories, and recent searches - but none of these behaviors were communicated to assistive technology. Keyboard users also had difficulty moving predictably through those suggestions.

Once users reached the results page, the filters relied on custom dropdowns and dynamically updated results. Some interactions behaved differently depending on whether a user clicked, tabbed, or used arrow keys.

A user going to the search section meant that they knew exactly what they wanted already, yet the accessible experience prevented them from getting what they wanted. The barrier was sitting directly between purchase intent and product discovery.

The Roadblocks

A few issues stood out during testing.

• Autocomplete suggestions appeared visually but were not consistently announced to screen reader users.
• Keyboard focus did not always move through search suggestions in a predictable order.
• Some suggestions could only be selected reliably with a mouse.
• The active suggestion was visually highlighted, but its state was not clearly communicated to assistive technologies.
• Filter controls used custom components that did not consistently expose their name, state, or available options.
• Expanding or collapsing filter groups sometimes caused unexpected focus changes.
• Applying a filter refreshed the product results without clearly informing users that the results had changed.
• Selected filters were easy to identify visually but harder to review or remove using a keyboard or screen reader.
• Some range sliders provided very little context about the minimum, maximum, or currently selected values.
• When no products matched the selected filters, users were given limited guidance on how to recover.

All of these combined made the shopping task slower and difficult to complete independently.

Why It Mattered

Search is one of the highest-intent parts of an e-commerce experience. A shopper who searches for a specific product, size, category, or feature is often further along in their decision-making than someone simply browsing the homepage.

If that customer cannot move through autocomplete suggestions or understand the available filters, the problem is not just an accessibility issue at that point. The client was losing high intent customers.

How We Approached It

We looked at search as a complete journey rather than testing the search field in isolation.

The flow included:

Enter a query → Review suggestions → Choose a result or category → Refine results → Review products → Adjust filters → Reach a product

We tested that journey using keyboard navigation, screen readers, zoom, and different interaction patterns.

For autocomplete, we looked at whether users could understand when suggestions appeared, how many options were available, which suggestion was currently active, and how to make a selection without using a mouse.

For filters, we reviewed whether each control had a clear label, communicated its exact properties to assistive technology and were clear to navigate.

We also looked at what happened when things did not go perfectly.

Could users remove a filter easily?

Could they recover from zero results?

Could they change a search term without having to restart the entire journey?

Could they understand how many products remained after refining the results?

This helped us identify where the experience needed to become simpler and more predictable rather than simply adding accessibility attributes to the existing interface. 

What Changed for the Business

The improvements also changed how users moved through the shopping journey.

In the weeks following the updated search experience, the client saw:

• 24% fewer abandoned searches
• 19% more users moving from search results to product detail pages
• 26% more search-led sessions resulting in an add-to-cart action

The shoppers who already showed purchase intent were now more likely to reach a relevant product instead of dropping out during search or filtering.

What We Learned

The search experience includes everything that happens after the first character is entered. For e-commerce teams, search interactions sit very close to conversion. A shopper may already know what they want. Accessibility determines whether the interface helps them find it or makes them work harder than they should.

In this case, improving accessibility did not require changing what search was designed to do. It required making the same product discovery experience work consistently for more people.

The result was better product discovery, fewer abandoned searches, and more customers making it from intent to product.