the search for a better search
Every good project starts with the search bar: we find files, ask for answers, and browse for inspiration. If searching is the key to success, then why aren’t more Dropboxers using it?
For this project, I was the lead product and content designer working on the Dropbox web surface.
the design process
the design process
define the problem
Feature utilization analysis showed that search usage across web, mobile, and desktop surfaces drives a 3.5% increase in paid usage retention but only 7% of web users are actively searching. How might we increase the number of Search Weekly Active Users (WAUs) and improve overall retention?
discover the users
persona explorations
Using existing user research on search behavior and performance metrics, I developed three distinct user personas that embodied the spectrum of individuals we wanted to engage in the search experience: the navigator, hybrid, and searcher.
experience journeys
After developing the personas, I mapped out a unique journey for each persona and hosted a brainstorm with cross-functional team members to identify solutions in the context of real user experiences.This resulted in over 35+ new ideas to explore.
develop ideas
idea prompt suggestions
In user research, we often heard “I wish I could search by describing what something looks like.” This functionality already exists but users didn’t know about it. To improve discoverability of search capabilities, we could introduce a cycle of prompt suggestions into the default state of the search bar.
idea query refinement
Sometimes, the big idea comes from doing the basics (or doing what you should have done in the first place). In the brainstorm process, we realized that we weren’t offering an auto-complete experience that aligned with our user’s expectations. Auto-complete is table stakes, and we needed to get on board.
idea clean zero state
A “zero state” is necessary from an engineering perspective, but our current zero state provides a lot of clutter. What belongs on a zero state and what doesn’t? We can explore cleaning up this experience.
idea new search variables
People search in the context of their workflow. If we can identify new essential variables (recently opened, frequent collaborators, high activity), we can refine searches to provide high quality results.
deliver solutions
After scoping and stack-ranking our ideas, we decided to start with the basics and build from there. I created a search bar that surfaces search ideas based on user behavior, a system to surface accurate query predictions, and a long-term strategy for future search improvements.