Testing Antipattern: Focusing on The Business

Some people say testers should focus on the business and its interests in striving to be better, faster, and cheaper. Often, the folks who say this aren't testers anymore, just peek at their resumes. They try to somehow map the work of testing to improving the bottom line directly through cost savings or delivering features faster. But, this is in direct conflict with the role of the tester. Testers do slow down software delivery. Testers do cost money. Testers do annoy other disciplines who just want to go fast. That’s kind of the point!

Making sure software is of high quality means that quality should gate the release of the software, and testers are specialists in assessing the risk in software quality and poor user experiences. Their job is to protect the consumer and the company from software manufacturing problems. Their job is not to speed up the assembly line, worry about cost cutting, or coddle management whose bonuses are short-term focused on raw production rates. Their job certainly promotes the interests of the business, but only indirectly.

Great testers should embrace the fact that they are a ‘cost center’ and that their job truly still is to defend the customer and the business from itself. Do you really want the tester of your autonomous car, missile, spacecraft, operating system, or even your email app to be aligned with ‘management’? Or do you want to simply forego testing for any of these applications? As a consumer of software, you are no doubt in favor of testers doing their job, to represent and defend your interests as much as possible. Even if frustrating in the near term, the mid-to-long-term business strategy needs these testers, as it will keep the business healthy with a quality-first approach. Companies with dedicated testers and quality engineers that focus solely and uncompromisingly on the quality of the product are the companies that win and retain markets.

Modern and “agile” testing efforts seem great in the near term. It is easier, more fun, and the projects do move faster without dedicated testers. But as software becomes more important to society, software testing and engineering should actually behave a lot more like NASA and Microsoft of old.

“It’s an ideal approach in private industry, where you want to be first to market. You get your product out there and then you release updates,” said NASA SA Technical Fellow Martha Wetherholt. “But when you’re heading for one launch of one vehicle, where safety needs to be proved and documented, a straight agile approach may not be the best option.”

It is a real problem that testers have limited career growth in modern software companies, and much of that motivates senior testers to gravitate toward the business roles, the product manager roles, or the developer roles. Rather than accept this trend, we should work to change this promotion ceiling for testers. If we don’t break the career ceiling for testers, the best testers will continue to leave the field, and the ones that remain will continue to align with the business just when they are at their best.

Microsoft of today is a poster child for the business alignment of testing. It used to be the shining city on the hill in terms of software testing efforts and delivering amazingly innovative and complex products to the masses and enterprises at the same time. Today, most dedicated testing jobs have been eliminated. Most testing at Microsoft is supposed to be done by junior engineers, but it really doesn’t get done, evidenced by a recent open letter from the CTO’s of major companies citing their dissatisfaction with Microsoft Windows quality: 

“It appears that there is a breakdown in the testing process. The Windows 10 insider process is not able to identify issues on released products. When your own products break with these releases, it is clear that current testing processes are not good enough…..

We want Microsoft software to be such that we can indeed install all updates and patches immediately without reservation. As it stands right now, we do not trust the software and the patching quality enough to do so.

I thank you in advance for the opportunity to share with you your customers’ views.

Susan Bradley

Moderator at Patchmanagement.org”

Ouch.

Where did all those great Microsoft testers of the past go? Microsoft used to employ tens of thousands of testers. On critical engineering teams, there used to be more testers than developers. These testers moved up the ranks, but there was a glass ceiling for testers versus the developer and product management tracks--so the great testers changed career paths to get a few more promotions in before they retired. Those testers that remained were either politely let go or magically turned into ‘engineers’ who focused on basic testing but might occasionally add some poor code to the product as well, to get promoted in their new role. A few left Microsoft as they saw this wave coming, but they too grew weary of the lack of career growth in their newfound companies and often turned to development or management roles. Worse, some of these respected testers will ‘brag’ about how they are no longer a tester and discourage new talent from entering the field. Worse still, they often repeat the process at their new companies and slowly destroy dedicated testing teams as they see it as a way to personal promotion. The diaspora of arguably the world's greatest testing force, have either faded into the background of engineering or continue the practice of destroying dedicated testing efforts wherever they go.

Too often late-career, and end-of-market companies gravitate toward avoiding real testing. This is dangerous for careers and companies and for our new software-powered world. That said, natural selection is an amazing force in business, and those who focus on testing and quality will win in the end.

--Jason Arbon, CEO, and still Tester, at test.ai