Human Condition Safety (HCS) was a startup focused on construction site safety. They were building a platform that would combine virtual reality training scenarios and IoT devices for on-site safety to lower risk on construction sites and in industrial facilities.
Their initial project with us pre-dated my time at ECP. About a year later they came back with new funding, ready to continue building their platform with our team. We committed a team of 5 software developers and one product designer to the project.
After a few months we realized they needed strategic leadership on the project. At first, we tried to solve this by having the designer pick up product management tasks, but eventually we realized that we needed another person with more experience added to the project. I was given this task. My contributions to the project were to conduct day-to-day product management tasks, review work for QA purposes, mentor the designer on the project, and tackle any design tasks that she couldn't fit into a typical day.
Challenge 1: Incomplete Design, Overworked Team
Before coming back to us, HCS had hired an outside UI designer on contract to design the native app for worker interaction with the platform. Unfortunately this designer was focused purely on visuals and the results were an incomplete system. Mostly, we got a few key screens to work with. Before I came on to the project the team was doing their best to work around this, but with the amount of work to be done and their relative inexperience it was hard to figure out how to move forward with incomplete work.
Day-to-day management was turned over to me. This allowed the designer to focus on creating good UI work and a complete design system. The developers got more efficient QA and were creating less bugs. I was also able to act as an intermediary between them and the client to improve productivity. I also streamlined our process to remove redundant communication and make clearer what work was coming up.
After 4 or 5 hectic weeks, the team began to ship well-tested code with polished visual components. We improved our velocity and were able to put more attention into prioritization and planning new features. Customer discovery became possible again to make sure we were only building necessary functionality.
Challenge 2: Expanding Functionality
Once the MVP was pushed live after my first 5 weeks on the project we were able to focus on new feature development more effectively. The original designer worked on making improvements across the MVP screens and states, and tackled new functionality as time allowed. I was able to change much of my product management time to focus on new feature strategy and design as well.
As a result, he team was able to ship several new features much faster than previous development efforts had allowed for, at a much higher level of quality.
Improved Password Creation
The original account creation password had just included a field for the user to put their password into, with no guidance on the rules. That field masked the password and then you had to repeat it in a separate field. Unsurprisingly, this led to a lot of frustration during the registration process. We needed to streamline it so people wouldn't have a poor first experience.
We ended up testing and implementing the following screens. Now users only had to enter into 1 field, and they could see what they were typing. All the password rules were visible up front, and the user got confirmation as they passed each requirement. The button to submit didn't become active until these requirements were met.
Once IoT devices were ready for testing we needed users to be able to provision devices to their user profile. These devices were small bluetooth tags they'd wear on their vests. At the end of each day they were dropped off to charge in base stations. Next day they'd have to grab a tag and associate it with them again. This process had to be very simple and clear. There was no room for technical jargon, so we took great pains to make the language plain and simple.
The third feature that I helped to streamline was requesting location permissions on iOS devices. Originally, the feature just popped up a request for location access without context after the user installed and opened the app. I reconfigured the feature to prime users to grant permissions instead of deny them. Permission wasn't requested until a feature requiring it was triggered, and first the user got a screen explaining why it was needed. If a user denied it they couldn't use the feature, but trying to use it would trigger instructions on how to enable the necessary permissions.
This project presented a great number of obstacles. Ultimately, we were able to stabilize it through ruthless prioritization, actively working to improve communication, and making smart choices about when features were good enough vs needing more effort put into them. HCS was acquired by AIG Insurance at the end of 2017 for their platform technology.