Defining Productivity @ ActivTrak
ActivTrak is a workforce analytics platforms that allows users to see insights on how they can optimize their work day and create healthy work habits
Role: Design Lead, Core App Team
Responsibilities: User Interviews, Prototyping, Usability Testing
Timeline: Q3, Q4 2021
Status: Released
Background
Activity classification is an integral part of the ActivTrak core app experience. It allows users to categorize activity data and classify it as productive or unproductive for their team. This in turn affects how users see our data in the platform and the insights provided for their teams. For example, a marketing team may find sites like Twitter or Instagram productive to their role.
Research and Exploration
After creating a survey and sending it to all admin roles in our platform, I was able to identify and interview admins who use the Classifications feature regularly. I worked with the Product Manager to set up user interviews and gain insights into how these users were using classifications, how often they use it and how long it takes them to complete their tasks.
Framing The Problem
Legacy design below:
75% of admins in our app rated Classifications as one of the most difficult and time consuming features, but it was also rated as the most used. On average, it took admins 2-4 hours a day to classify activities. I created a user flow to document their workflow and find areas of improvement. From their I was able to create our HMW statement.
How might we enable users to quickly classify and manage activities?
Kickoff Meeting
During the kickoff meeting, engineering voiced concerned on development issues.The old framework that we were using was going to hinder our expansion of the platform as we continued to grow and create more robust features. At this time we decided to start planning out a process to build new features in a new framework (React) while slowly chipping away at tech debt which comprised of all of our old framework (Angular). Because Classifications was one of the first features with the most complications in Angular, we decided it was a great candidate to rebuild using React.
Designing with Development Considerations
With the transition to React, I had to do my own research on which components would be best suited for this feature and with the help of engineering, estimate a realistic design and build time that would still leave us hitting our quarterly goal of delivering a new feature. I used Material designs component repository to find a table that allowed for us to display activity data and have built in bulk actions to help users complete tasks quickly.
Usability Testing & MVP Requirements
I conducted unmoderated usability tests via PlaybookUX to validate my design decisions. After a completing a successful round of usability testing and grooming meetings, I worked with the PM to determine a proper MVP and release schedule for this feature. Because this was a major feature, we wanted to minimize the risk of bugs and account for change management. We decided to slow down our release cadence by releasing to small sets of users at a time.
First, we recruited the top 10 users of our feature. During this time I created a trigger in the app that would prompt users to leave feedback and rate the new feature via Intercom. During this phase we identified a few bugs that were hindering usability. Once these were fixed, we released to 10 percent of our users. We monitored feedback for 3 weeks and continued to fix bugs as they were identified
Release and the Future
Unfortunately we delivered this feature a quarter and a half late. This was due to improper scoping of the technical implications that come with changing frameworks. As an advocate for the user I had to make sure I was reminding the team of the goals and what was absolutely necessary for MVP.
Currently we are still collecting early access feedback from users. With the interface being completely different from the previous, some users noted they did not like it simply because it was different. Other users praised that although it was different, they really appreciated how easy it was to learn and how simple we made a task that was previously time consuming and confusing. If all goes well, we will release to our entire customer base early in Q1 2022.