The objective of this project was to implement a new search system that enhances the user experience and enables Shelter users to more effectively filter content.
To initiate the project, we conducted a moderated usability study with 5 participants, focusing on our internal search, to assess its performance. We felt this was particularly important after previous quantitative research indicated that 0 result searches contributed to an increase in calls to our helpline.
Next, we analysed analytic data to identify the most popular search content among users. This analysis helped us establish benchmarks for the functionality of our internal search and measure any changes made. During this work, we also focused on examining pain points and areas where users experienced difficulties or dropped off during the search process.
Based on the insights gained during the discovery stage, the following improvements were proposed:
.png)
We prioritised these tasks and divided them into two phases. The first phase focused on tasks requiring minimal design work and followed best practices, such as improving spelling suggestions and search result ordering. These recommendations were passed on to the development team for deployment while we worked on phase 2.
For the second phase, we conducted an ideation workshop to address the remaining tasks and determine the best solutions. During this workshop, we created paper wireframes to explore different design options and iterated on the concepts that received the highest votes.
Using Figma, we developed high-fidelity mockups based on successful wireframe iterations. These mockups aimed to visually represent the improved search system. The design was then handed over to the development team for implementation in a staging environment, allowing us to test the internal search as it would appear and function on the live site.
To validate the effectiveness of the new search system, we conducted usability testing with a representative group of 5 users on both mobile and desktop devices. The testing focused on evaluating the usability and user satisfaction with the new search filters, misspelling handling, auto-completion, and overall improvement of search results.
The results of the second usability study showed that 9 out of 10 participants were able to complete tasks more quickly than before and more frequently landed on the correct pages compared to the first study. Furthermore, the majority of participants were more likely to recommend Shelter after the second study. Based on these findings, we concluded that it was appropriate to launch the new designs.
Following the implementation of the new search system, we observed the following improvements compared to the benchmarks established during the discovery phase:
These results indicate that the implemented search system successfully enhanced the user experience, providing more relevant results and reducing user frustration while navigating the Shelter website.