Solutions Engineering / Ecommerce UX
Diagnosing a Product Search Problem Through User Behavior
A recurring "bad search" complaint turned out to be a visual hierarchy problem rather than a search-relevance problem.
Overview
Sales staff began reporting a recurring pattern: customers searching the ecommerce site by SKU were frequently selecting the wrong product from the results. The confusion sometimes became apparent only after the customer began discussing the product with a salesperson.
The Challenge
The obvious assumption was that the site's search functionality was returning or ranking products incorrectly.
When I tested the search behavior, however, the correct product was consistently being ranked as the closest match.
The problem was somewhere else.
What I Noticed
The results were displayed in a two-column grid.
The highest-ranked result appeared in the upper-left position. The second-ranked result appeared immediately to its right, placing it visually closest to the search field and interaction area.
Sales staff had noticed that customers were disproportionately choosing that second result.
This suggested that the search relevance itself was working correctly, but the visual presentation was influencing which result users perceived as the best match.
How I Tested the Idea
I located the search-results presentation code in the site's source files and used the browser and DevTools to experiment with the layout.
I adjusted presentation, spacing, and result organization while testing the changes directly in the browser.
Rather than replacing or rewriting the search system, I focused on making the visual hierarchy accurately reflect the relevance hierarchy.
Visual Reconstruction
Simplified reconstruction of the search-results pattern
Two-column results
The second result sits closest to the search interaction.
Relevance-led results
The order now reads clearly from closest to least-close match.
The Solution
I changed the search-results presentation to make the closest match visually unambiguous and to create a clearer progression through subsequent results.
The revised layout made it easier for users to distinguish the primary result from secondary matches.
The Result
After the revised layout was deployed, the sales team responded positively and thanked me for addressing the issue.
I have not received further reports of the recurring pattern of customers selecting the wrong search result.
What This Demonstrates
- Technical Troubleshooting
- User Behavior Analysis
- UX Problem Solving
- Root-Cause Analysis
- Frontend Development
- Ecommerce
- Solutions Thinking