By Janet Weddle, Technical QA Manager at Keywords Studios
Technology is evolving at pace; you see new games and apps released every day that need detailed testing to provide the best player experience.
On the surface, UI Automation seems straightforward. So, why do your hear so many stories about terrible failures?
Long ago, before my time at Keywords, I started out as a developer. The companies I worked for were busy and, usually, if you wanted something done, you had to do it yourself.
If I saw a need I would go to my peers or my managers and say, “We really need this, it will save us time…money…or reduce errors”. And they would say, “That’s a great idea! You should do that!”.
Most of the time these were small tools created for my team or myself because I was the designer, the PM, the IT department, the developer, the automation engineer, the tester and the user.
Then my role changed. I became a manager.
Because of my previous experience with automation, I found myself managing an automation team made up of highly skilled engineers.
We worked on different frameworks and different projects. And I started to see some of the same roadblocks come up over and over.
Imaging the following scenario…
Someone you know has been invited to a leadership meeting. Several hours later, they come to you and say, “We need to incorporate automation into our pipeline”.
You wait for more and then reply, “Okay”. They explain that automation is faster and less error prone and our product should take advantage of that.
You know what automation is but what you don’t know is what they’re looking for. What kind of automation and where in the pipeline?
When someone comes to you asking for ‘automation’ there is often a pause as if you are going to jump in with answers, to questions that haven’t been asked.
Since I’m from Seattle I’ll use a Seattle reference. Going to your team and saying, “we need automation” is like me going to Starbucks and saying, “I need coffee”.
We’re in the right place and asking for the right thing but there are so many flavours and types that you will never get exactly what you want unless you ask for it. You need to be more specific.
Usually, not always, the first part of a successful automation plan isn’t hiring automation engineers. It’s looking at your pipeline. It’s looking at things as they are now.
Look for bottlenecks. Look for things that you do repeatedly or things that take a long time. Start high level and even draw a simple chart (like this one) that helps you determine where you should start your planning.
Once you have a direction, dig deeper.
In a recent digital event, I shared these and other key steps that can help you change your approach to UI automation so you can make your next story an awesome success.
You can watch the on-demand event on the Keywords website to hear about the Functionality QA UI test automation best practices to help raise your game.
Key takeaways include:
- Learning from the past – why automation may have failed before
- Why you should reduce your automation focus
- Specific items to review before starting an automation effort