The KPI (Key Performance Indicators) for customer driven software are pretty easy to come by, ranging anywhere from increased sales to better customer reviews. When designing the UX for internal employee tools, however, it seems to fall into one of two goals: reducing time and errors per transaction. (Once a process has been created and codified, user (employee) satisfaction is hard to fight for – I talk about this in another post.)
In this case I’m going to cover the importance of research to find the right goals. I was asked to join a development group as they started on developing a new redesign of an old product. The product was being redesigned because it had been running in old database software and was part of a larger push to not need to renew the license. As with most internal projects it was getting started behind schedule and I got pushback for requesting user research. They already had all the specs they needed from the old product, big red warning flag.
I was able to justify research to collect the current task time and error rate. While collecting this data I also made sure to do contextual analysis to see if the current system was meeting their needs. In other words: what were they were using outside the system? The number of papers, post-it notes, emails and excel work sheets was another big red warning flag people were working around the software, instead of using it. On further analysis it was obvious that most of the paperwork had to do with each user’s personal stash of venders they were buying from with nothing in the system for users to compare prices amongst those vendors.
This would be some new features for sure. To justify the time and work needed for it, I collected the average dollar amount per order per employee and showed there was a discrepancy in the tens of thousands of dollars.
While designing the new system I built in some gamification to encourage the users to give up their personal vender knowledge. A leaderboard was created, but in the first round of user testing the same most senior people kept at the top with little change for creating the least expensive invoices. I changed the scoring to also give points for the number of vendors entered into the system and it changed everything. Now users could compare what the vendors they had with the vendors in the system. If they were cheaper they would only get the points by adding the vendor to system.
The extra features added an extra sprint to project but first month the new system was in place it paid for the license extension needed for the old database software, and every month after that was pure savings which kept increasing as time went on.