Lessons on Agile and Crowdsourced QA from the Dial-Up Days | Inflectra

Lessons on Agile and Crowdsourced QA from the Dial-Up Days

This article discusses some of the earliest attempts at crowdsourced QA and extreme agile development that predate the agile manifesto and where they went wrong.

Remember the old Dial-Up Days?

Today I was cleaning out an old filing cabinet, you know, the cabinet we all have that is full of hanging files that we have not touched in 10+ years. It had to be done. Well, I came across an archaeological find. Nothing as good as a dinosaur bone, or anything cool like doubloons. It was a "Get Online" disk from the 90's. Yeah, now we use them as coffee coasters, they used to come in the mail every day, and there is no real value to them. But there is a lesson...

In the old days, these proprietary services were the internet. Sure, there were those of us who had other dial up connections, and low numbered CompuServe accounts (big shout out to Columbus, OH), but for the masses, it was one of these dial up guys or nothing. The internet today is the legacy of proprietary internet dial up providers, you could argue for others like BBS, but the proprietary guys brought it to the masses. They also had their development teams keep floppy disk makers in Ferraris far longer than they should have been.

The QA Lessons from Those Days?

You see, the mindset in those days was, don't do quality assurance (QA), instead, let's croudsource it. BTW, crowdsource is a modern word, in a historical context. Back then it was a given that software was unreliable, and we all expected there to be bugs. I never worked for any of them, but I had friends that did, and the engineering teams knew that what they were releasing was not really ready for release. Each week, there was a new release to fix a few bugs, and another million disks sent to the world either on a magazine, of direct.

Guess what, it worked! For a while at least. Constant release and no/minimal QA, did eventually lead to better software, but at what cost:

  • Perception - having one of their email accounts is not a sign of expertise - there is a long tail to the memory of the bad old days, and having an email account that ends with one of their domains is not a way to engender people with your superior technical knowledge

  • They made a fortune - One even bought one of the largest media companies in the world!

  • Business faltered - as new players started approaching the masses (Cable/Broadband/etc...), customers started falling off. You could now get reliable online access without the bugs (or at least not as many). Most had to change the business model from subscription to advertising based.

So what can we learn from this. Well, they were actually on the bleeding edge of several things... Crowdsourcing, they were doing it in a massive way before anyone else had thought of it. Agile, they were doing agile and iterative improvement before those words were coined. And that perception is a key element to long term survival.

Sure, you can take advantage of your customers for a while, and rely on them to act as your testing organization, but it will fall eventually. Then you are in a hole and have to fight your way out. Had there been a structured QA program in place, we might still have seen them make a ton of money, and they might be the organization deciding how we use the Internet to this day.

agile crowdsource quality assurance qa