Last week I was talking about trends in software testing with my friend and colleague, Griffin Jones. One area both of us are interested in is the downward pressure on test costs, which leads executives to outsource, offshore, or use a specialized service like
uTest.com.
In general, I’m in favor of the invisible hand of the free market. If someone can find a more efficient way to offer the same service for lower cost, well, bully for them. In the end, the competition leads to better, cheaper prices for the end customer.
The problem comes when the customer can’t tell the services apart, or at least can’t tell the good from the bad. In economic terms, this creates a lemon market.
In other words: If you can’t tell the good from the bad, it seems reasonable to at least buy cheap.
Griffin took this situation and added one simple question:
“Are we [the test industry leaders] just selling the best buggy whips in the age of the automobile?”
It’s a reasonable question, and I have a few potential answers.
Let’s consider the possibility that we are selling buggy-whips. Software Testing, after all, is work that anyone can do, and is best done cheaply, or, if can pull it off, you might be able to chase the “facebook effect” and get your customers to test for you for free.
If that’s true, then perhaps, as the old saying goes, we should enjoy our nice testing jobs while we have them.
There’s even a name for this trend: The “
death of quality.”
The question is: Can all software be tested and deployed in this way?
I mean, social media and websites and games, perhaps, but is your credit card company going to roll out features and watch for failures? How about eBay, or Amazon, or any company managing financial transactions? How about companies that manage personal information like Social Security Numbers, dates of birth, mothers maiden name? What about weapons systems, payroll systems, mission and life critical systems?
I’m willing to make a bet that companies that choose to manage testing as a lemon market, that choose to manage entirely to cost, are going to fail in the marketplace. We’ll see software that isn’t fit for use, software that doesn’t work, software that no one can figure out.
And, sooner or later, someone would get the crazy idea that they could compete by building software that is actually, well ... better.
One such company is Apple Computer, but it’s not the only one -- another is a small company in Seattle Washington called the Omni Group, that makes smaller tools, utilities, and games for emerging systems. They’ve been doing that since the emerging system was this little one-piece machine created by Steve Jobs called the Macintosh.
This leads me to a different question: Is the market big enough that we testers can find roles at these ‘different’ companies, these companies that care about software testing, and are propelled by market forces to have extremely strong software quality on the first release?
I suspect that depends on the size of the software testing market.
And yes, there is a real software testing market. Today there are magazines about software testing, conferences for it, companies that specialize in doing testing, outsourcing testing, test augmentation and test training. It is a real market.
And it’s a growing market.
Consider the past past twenty years, where software has injected itself into every aspect of our lives, from our cell phones to our social lives, our automobiles, music players and even our cameras. The folks at Xerox PARC had a name called it “
Ubiquitous Computing.”
If this trend continues, we’ll see more software in our lives, not less of it. Yes, yes, some companies will be able to delivery poor software, sense feedback from the market and respond.
But there’s this great thing called the free market.
Over the past few years, Apple computer grew from a scrappy upstart (that arguably existed so Microsoft could legitimately claim to have competition) to become a respected, leading-edge, radically profitable company that is the single most valued company in the world by investors. Yes, literally most valued: Apple leads the Fortune 500 (link) as largest market capitalization -- most valued because that is where investors choose to place their money.
I actually believe in the stuff we are selling; that the companies that choose to invest in skilled, thinking testing will win in the free market, the same way that Apple Computer invests in thinking and engineering.
I’ve made my bet, but I have to agree with Griffin when he points out that many companies
will pursue lemon-market strategies. That might even be good business sense for some of them -- but those strategies could distort the market and create unrealistic expectations.
The next five or ten years could be very interesting for the dev/test industry. Worldviews are going to be colliding, and we'll be battling in the court of public opinion.
I've made my bet, and you'll keep hearing my ideas here, in person, and other places on the intarwebs.
May the best strategy win.