Advanced Search
Overview
The Jackson Laboratory is a leader in biomedical research and mouse model development. Buying mouse models can be a complicated process for even the most seasoned scientist.
The Jackson Laboratory Mouse search aims to connect scientists with the right model by displaying information in an accessible and clear format for all user types. I joined the product team responsible for B2B sales and helped lead the redesign of their Advanced Search feature.
Role
Lead UX Designer
User Researcher
Prototyper
Timeline
June 2021 - ongoing
The problem
Customer feedback surveys provided insights that customers were trying to search for mouse models by a specific gene, but the search results were not providing accurate results. After searching for genes myself, it became apparent that the search was using keywords and did not verify the gene category on the mouse model datasheets.
Design Challenge
How might we make the mouse search more accurate and relevant to customer needs?
User Research
To kick off the project, I began to dig deeper into how customers currently search by looking through Pendo (our data analytics tool) to gather more information on how customers search.
I evaluated the search attributes that users had used between May 2021 - July 2021.
The most popular search category for users were genes and alleles.
According to funnels, customers tried multiple keyword variations.
When multiple keywords were entered the search engine did not return relevant results for all keywords.
Because the search results were not relevant many customers left the website without clicking on a datasheet.
Based on the findings from the search data and reviewing heat maps through Pendo Analytics, I was able to put together mindsets for how our users search for mice models. Two mindsets would benefit the most from an improved search engine.
Quick Searcher
This user glances at the search results, but won’t go past the first couple of search results. This person expects the search engine to be properly optimized.
Targeted Searcher
This user knows exactly what they are looking for and will search by a gene or disease area. This person will spend time finding the right model, but expects the results to have what they are looking for.
Dev Team Feasibility Check
The next step was to understand the technology and data structure constraints of the project. I was able to learn that the current search engine would not be able to recognize a gene, allele, or therapeutic area which are the top ways that users search. It would take too many resources and time to create the right connections to place every search term type into a category.
User Flows
After I understood the business and technical requirements, I began to map out all of the use cases for the advanced search feature. I began mapping out how the advanced search feature would adapt based on if the researcher was searching for a gene, allele, or therapeutic area.
Design Session
To generate a variety of solutions, I led a brainstorming session with UX Designers, Product Managers, and the Director of IT to help find relevant inspiration and wireframe possible solutions.
Card Sorting
Once I began putting together the user interface, I wanted to connect with the Technical Information Scientists to make sure I was connecting the filters with the right advanced search category.
Prototype (Iteration #1)
After I wrapped up the initial discovery work and worked with stakeholders to ensure the information was technically correct, I started working on prototypes. This initial prototype was made to show key stakeholders and help facilitate discussions around the feature.
Prototype (Iteration #3)
Based on conversations with stakeholders and engineers, I was able to iterate on the design and create a prototype for moderated and unmoderated usability testing. In order to keep the design agile and not take up to many of the dev team’s effort points, the idea was concentrated down to just a genetic search.
Usability Testing
I am currently running an unmoderated user test through Maze and have moderated usability interviews scheduled with current customers. My goal for the unmoderated test is to collect a significant amount of quantitative data to determine if enough users want this concept, while the moderated testing will be used to gather insights on how to improve usability.
Next Steps & Lessons Learned
The next step is to complete the moderated and unmoderated testing sessions in order to synthesize the data in an affinity map. I want to provide the stakeholders with validation that users want this advanced search and gather insights on how to improve the experience.
During this project, I learned how important it is to communicate across all of the design and development teams. Every check-in with the development team and stakeholders provided another necessary piece of the puzzle and help to inform the design process.