Positive Stories About Test Automation

I saw a post on LinkedIn today from @lgibbs that I thought would be excellent to share here too.

Link to the original post

Which says

I see so many posts about the negatives of test automation, and I get that. We need to raise awareness about what it can and can’t achieve. There are a lot of unrealistic expectations out there.

For a change, I’d like to hear some positive stories about test automation. Anyone able to share?

4 Likes

After two years of struggling to get automation done by my own steam at a startup, I moved to a new company that already has a mature automation framework and tbh it is so nice. My manager is super passionate about reusability, maintenance and readability/simplicity so it’s really cool. A lot of the time I’m trying to write a method for something, and I realise there’s already one made with the flexibility and appropriate wording for what I want, so I just like… use it… and then it works, and I don’t have to do anything because it’s all been set up in such a nice way.
I do complain a lot about it as well, obvs, because I haven’t done Java/Cucumber/Selenium before, or worked in such a massive automation project before. tbh though I love so many things about our automation suite and I’m happy to be somewhere where there are lots of people who can and want to help write it!

2 Likes

My first experience with automation was a bit scary, I was working less than half of a sprint on a automation project. I had two other projects I was also working on so I could not focus properly. The automation framework was done in Java, using Selenium, Jenkins and TestNG. It was a big framework and it seems complicated to me (at the time).

Few years latter, I got tasked with making and API automation test framework and I decided to go with Java and as it turns out, after learning more about coding and automation for several years, everything made sense now and there was no place for fear.

1 Like

We had a 10.5 days regression set (including test data creation (1d = 8 hours)) and this should have been tested sometimes even multiple times a week, which isn’t possible with the amount of tester-capacity we had. They never heard of test automation before, so when I joined the project I suggested it… they didn’t understand it and didn’t want to invest in it.

So I automated all those regression tests & test data creation steps in my free time just to prove them a point. And in 1 year of time we made 483 days profit since we used the automated test suite to check for regression. It even caught multiple bugs which wouldn’t have been found by manual testing since we couldn’t always finish manual testing due to lack of time.

2 Likes

As one of the rare coder-turned tester types, I have enough horror stories of test automation failure in my early days. If only I had been given a copy of the giant Dorothy Graham, Mark Fewster: experiences of Test automation studies book, I would have avoided a few. But no, it does turn the corner. For me it was all about writing our own test frameworks, and never fighting with out-the-box stuff, and we have always succeeded. Mainly in creating CI/CD pipelines that scale up well and become so much a part of the SDLC that release cycles come down and down. The goal is faster releases, not fancier or cooler tooling. That’s how to win by small amounts, win just a little each day. Small wins.

In my last 4 or 5 test jobs I have always enjoyed most seeing the time to run a full regression suite reduced to the point that devs can throw code into the branch quicker. That’s my joy source, sometimes it means killing a few test cases, but cranking that handle faster rocks.

1 Like