Redesigning INSPIRE
Redesign of the scientific information platform INSPIRE used by particle physicists around the world. Currently live under inspirehep.net.
Presented user-centric approach in International Conference on Computing in High-Energy and Nuclear Physics (CHEP) (paper).
Increased Net Promoter Scope by 40%.
Improved 20% the task completion rate.
Role
Lead UX Interaction designer
Product Manager
Challenge
INSPIRE (fka SPIRES-HEP), the first website in North America, was a searchable database created at Stanford in 1991 to help physicists worldwide share research papers on particle physics. The platform had kept its 90s design and functionality for the last 30 years and it was time for a technology upgrade. The only challenge? Users loved it exactly as it was.
Physicists use INSPIRE every day for 15+ years and they were familiar with all the tips and quirks that are needed to use it. However, there were a lot of inefficiencies and improvements that could be made, and given the backend technology needed to be updated, we decided to redesign the frontend too.
How Might We…
redesign a product for thousands of loyal users with strong opinions?
Understanding the users
The main INSPIRE users can be described by the following two personas:
INSPIRE author profile page (before redesign)
Information architecture and mocks
I decided to do an audit to deeply understand how useful the current interface was.
I started with the Author profile page. There was no data to understand which of the sections in the author profile were used most by users.
I decided to recreate their profile page, this time in a low fidelity mock in Balsamiq, so that they don’t associate it with the current design.
I printed each section and asked them to create their own profile and the profile of another author (to see whether there would be a difference in the sections they selected).
After the exercise I collected valuable insights:
Users selected a very limited number of fields to be shown prominently (even though in the interview prior to the exercise they stated that all fields are important).
Their behaviour was different based on whose profile they were viewing: for their own profile, people wanted to stress out their career history and their achievements, whereas for others they would like to see a quick overview of the author’s research works.
Prototyping
Following the mocks exercise, I created the low fidelity prototypes of the author page and asked users to perform the tasks they perform regularly using the prototype: Find the author’s papers, Find the author’s citations etc.
During the prototyping phase, I discovered that some fields were missing from the mocks, (eg. their career history was important to them when they were looking for new opportunities) so I added them through iterations.
MVP design
70% of the users reported that they would use the MVP version permanently!
Activities
User research
UI Design
Information Architecture
Prototyping
Usability Testing
Tools
Figma
Balsamiq
Invision
Team
Final designs
Apart from the author page, I redesigned all INSPIRE pages too:
Final design
Author page
Following user feedback, the new redesigned author page contained only the information that users would need in a first glance.
Citation Summary
Citation summary is hidden under a toggle that only when the user needs, can enable.
Apart from the author page, I redesigned all the other INSPIRE pages too:
Literature search page
Through multiple iterations, I added elements based on the needs of the High Energy Physics community. For example:
Users needed a quick way to see when was the author searched more active, therefore a histogram under Date was added.
Due to large experiments, users wanted to filter out papers with more than 10 authors.
Users needed a quick way to extract a paper’s pdf or view the list of all papers that were citing this paper, so I added these.
While users may love the product, there might still be pain points or areas where the experience can be improved. Following the user interviews and the usability tests on the new paper prototypes, I discovered that users were very familiar with the current interface, but were missing features that would make their research more efficient.
While redesigning a product that users use daily and love, I learnt to:
Identify pain points and areas for improvement
Respect the past
Collect user feedback continuously
Communicate clearly and transparently
Identify what users love about the current product and why they use it. Instead of making drastic changes all at once, introduce changes gradually. Users are so familiar with the existing product, that I needed to make small steps along the way to make sure they are onboard with the changes.
Throughout the design and development process, I actively seeked user feedback via several channels (interviews, support tickets, data). This feedback was essential to help me iterate and refine the product further.
I wrote several INSPIRE blog posts to communicate any upcoming changes or new features to the user base. It was really important for the physics community to understand the rationale behind the changes and how they will benefit. This helped to manage expectations and build trust with the community.
Literature detail page
When users clicked on a paper, they got more details about it. This info was again selected based on the feedback from user interviews and usability testing with the mocks.