Are you a large company suffering from slow release cycles, and there’s a threat of losing out to faster competition? Or are you a startup focused on shipping valuable software quickly without much structure? Whoever you are, and whatever your size, you may be feeling the pinch of needing test automation in your workflow. But that can be a touchy subject. Many people are aware that an automation effort has a high upfront cost, and it can take awhile to see the ROI. In this article, I’ll lay out some tips on how you can persuade your boss to start down the automation path, along with how to be successful once you’re on it.
Return on Investment
The biggest thing on your boss’ mind will be: What’s the return on this effort? What are we going to gain, and how will this make us better? So for all the tips below, the answer to those questions should color everything you do about persuading them to start down this path.
“60% of the Time, it Works Every Time”
Your boss may be aware that automation at other companies didn’t work well, while at others, it did. It can appear mysterious and cause hesitation, until you can explain why that happened. Present case studies about those companies and about their patterns (or anti-patterns), so that you can avoid making the same mistakes they did.
Set Clear Expectations
Many teams, along with their bosses, go into this effort intending to automate everything. I’m here to tell you: don’t do that. It’s the quickest path to failure. Very few things actually need to be automated, and, oddly enough, the less of it you have, the more valuable it becomes. Fewer moving parts equals more stability. As such, you’ll need to set a clear expectation of what will be automated, what won’t, and why.
People love pictures, especially when they convey important information. Getting metrics from automation gets overlooked a lot, and it’s a shame because this is what can make or break the effort. There are many tools or plugins available that help display data about what tests passed or failed, what categories of tests were run and what parts of the system might have problems. Scope out which ones are relevant to your effort, and write your reporting to play well with your choice of reporting solution.
Show, Don’t Tell
The most powerful example of persuasion will use just that: an example. Can you make a proof of concept to show your boss? Bonus points if it’s simple enough that a non-tester, non-developer can use it.
Once Bitten, Twice Shy
Some of you may already have tried automation and it didn’t end well. It can make the conversation more challenging, but not impossible. That’s why you need to arm yourself about why the previous automation effort failed, and present a plan for what can be done differently.
Don’t Ignore Maintenance
A novice automator can throw together a framework in a day… and then it’ll collapse under its own weight with regard to maintenance. It may take months or years, but it’ll happen, and then the effort will have to start over. A needy framework is like a many-headed Hydra. Be sure to consider how to make your framework easy to maintain, so that the work isn’t focused on feeding the beast, and so that you can focus on the correct thing: shipping quality software.
Broadcast What’s On Your Mind
The most important part of the sell is what the ROI will be. The next—arguably of the same importance—is communicating what you’re thinking, and what you’ll be aware of, as the automation effort continues. It not only shows you’re aware of the pitfalls and goals, but also shows that you have a plan instead of making the mistake of just jumping in and doing automation.
I wish you the best of luck in the art of persuasion—you, your boss, your team and the whole company will benefit from the automation effort! You are welcome to share your thoughts in the comment section below.