I used quantitative research methods to understand current workflows and establish performance benchmarks for a corporate product used to triage and process scientific research.
My Contribution
-
Scheduled remote usability sessions
-
Time-coded all session recordings
-
Analyzed data and made recommendations
-
Created redesigns based on observations
Project Overview
Problem
Client
Duration
Role
Collaborators
Output
Establish performance benchmarks for a corporate product used to triage and process scientific research
ProQuest
December 2018 - July 2019
Researcher, Designer
UX designer, UI designer,
developers, product manager
Metrics for product's core workflow
Sketches and mockups for UI redesigns
Project Phase | Lesson Learned
Discovery Phase
The Drug Safety Triager (DST) is used by pharmaceutical companies to assess and prioritize recent scientific research relevant to their drugs.
Each client has a unique instance of the DST product, which means that each version had different interactions and workflows that had to be accounted for.
I watched several hours of pre-recorded usability testing from an early DST study to get up to speed on how the product works and to better understand how the current designs support
(or complicate) user workflows.
Afterwards, I made a flow diagram that turned out to be a useful document both for me and for communicating with stakeholders and making sure that all of us were on the same page about the expected workflows to be studied ahead of the benchmarking phase.
Lesson Learned
Flow diagramming and object modeling are useful for aligning stakeholders expectations about complex workflows.
Analysis Phase
The first big challenge was finding common tasks to measure. Not only does each client instance operate differently, but each individual reference also has a unique workflow.
To make coding and the session tasks easier, I worked with the team to identify five high-level task rollups:
-
identifying country affiliation
-
searching
-
reading
-
getting full text
-
note-taking
With these rollups, we were able to deal effectively with the variations across references and between different instances of the product.
I was still able to return to the raw data if there were questions and to provide additional granular analysis about a particular part of the workflow on an as-needed basis.
2
4
8
114
instances
reviewers
hours
references
Lesson Learned
Complex systems can be dealt with by approaching them from different scales, both more granular and more general.
Ideation Phase
New designs had to take into account stakeholders’ reticence to change a system that’s already working, even if it isn’t working very efficiently.
The current DST has a vertical layout: lots of up and down scrolling that translates to lots of wasted time. I sketched several different layouts -- a 60/40 columns, 3 columns, and horizontal. In the end, the most effective use of space was 50/50 vertical split.
Reviewers need to reference additional content that lives in different locations and apps. Leaving the DST slowed users down and made for inefficient workflows. Incorporating
secondary tab navigation will allow users to quickly move between content and reduce the need to access resources external to the DST. When combined with accordion menus, we were able to condense a lot of content and actions into a small, economical space.
After a second round of sketches and mid-fidelity wireframes, I build an interactive prototype using Sketch and InVision.
Lesson Learned
There’s value in ideating from multiple perspectives, even when the constraints already seem set in stone.
Gallery
