

Discover more from Robert Roskam's Newsletter
Early in my career, I had an idea to save hundreds of engineering hours and enable the company to respond more quickly to market conditions.
What was the idea? We should automate integration testing.
I know. Not very amazing or interesting.
The status quo was each test case took 30 min, and we’d have to do at least 20 to be ok, but honestly, we’d rather have 80. And we’d have to do this after every material change, which we wanted to be weekly, but the math didn’t work out.
So easily 10 hours per week, and it could have been 40 hours. But we usually only did this once per month, because of how time intensive it was.
So I went off and came up with some automation over a period of about 6 weeks. I did my normal workload, and while I was doing the manual testing, I just added time to build up automation of it until I had the entire thing automated, and I could process about 100 test cases in about a minute.
I never asked. I just did it. If I asked, we would never have done it because people would have assumed it would have been a lot of work. Since I made it part of my normal work, nobody cared.
By the end, we had our vendors telling us to stop doing whatever it was we were doing in their test system because we were overloading the system.
Oh, and yes, I got that raise.