Imagine a library with 50 million books, but instead of books with accurate metadata from a publisher, the information is submitted by 3 million different people, and each person provided information using varied categories. To make things more interesting, your job as a librarian is to catalog the books for your 35 million patrons. That scenario maps pretty well to my time as lead product designer on search at Etsy. Trying to understand and organize those 50 million unique handmade or vintage items was what I was up to as a member of the search team for almost 3 years.
I have worked on search from a handful of different angles, but the main focus had been to create an understandable UI backed by qualitative and quantitative research, work with other parts of the company in getting more structured and accurate metadata from sellers while working with our taxonomists in organizing the inventory, and guide users to what they are looking for through filters and other means.
I started by updating the search results page to feel more like the other main pages on Etsy's site. After conducting some user research and testing, I also updated the UI of several search filters.
The next step was to make search more responsive by taking better advantage of different screen sizes. To make this happen, I worked very closely with engineers and together we rewrote all of the legacy HTML and CSS on the search page using Etsy's new CSS tool kit.
I spend a good chunk of my time working to understand the difference between low-, medium-, and high-intent queries and understand how we could create different and more tailored experiences in each case.
One example of a low-intent query is the below idea. The short summary is that when a user enters a query, we run a few calculations about the diversity of the result set between categories to determine the intent of the search. If it's proven to be low-intent, we show a group of selected subcategories that help the user dive in further to narrow down the search. You can read more in depth about this project here
We spent a lot of time thinking through different ways to allow users to filter (not shown here), and adding new filters when they became available, all while making sure we map to the most current visual language of Etsy.
I've also worked on the category pages for Etsy, finding and trying to streamline the other ways in which buyers can be connected with items in which they're interested.
We are continually updating the visual ui and running experiments to improve the experience. Search will never be done. The smallest details have made an impact on our metrics around search; things like font size, spacing, image size have all changed the experience.
I also worked on these features at the same time for all sizes of Android, iOS, and mobile web. Below you can see mobile web versions from when I started versus current iterations. I've spent a lot of time working on and updating the listing card along with other IA choices.
So there you have it, a quick overview of some of the areas of search that I have worked on. There is sooooo much more that I've worked on in regards to filtering, search intent, understanding queries, that I can't quite do it justice in this setting.