I was in Vegas recently for a conference, and decided to gamble. Just a little, of course, and I’d stop as soon as I was up. Maybe I’m too practical, but I see gambling as something you do with petty cash, not as a line item on the household budget. Not so surprisingly, I met my goal and stopped … but that wasn’t the interesting part of the evening.
After finishing my foray, I paused to watch two gamblers at the roulette table. Within seconds of spotting both parties, my immediate reaction was, “Hey, wait! Stop! Please! You very obviously do not have the financial reserves to be betting how much you’ve just plunked down.” Despite my forceful telepathic cries, they plunked away. “Don’t you know Vegas was built by losers?!” my mind shouted at volume 11 while they stood next to one another and bet on the opposite color. The older man bet red while the younger-yet-old-enough-to-know-better guy bet on black. This went on and on until I turned away, unable to bear the leeching of these poor souls.
And that got me to thinking: this scene has an ironic parallel to something from my automotive past. I’ve seen two companies—neither one with stable balance sheets—gamble a quarter of a billion dollars (yes, that’s right … $250M) on an interior electronics platform hoping that it would sell more cars. They probably wouldn’t agree with that number since: A) they haven’t truly added up all of the engineering hours, executive meetings, supplier purchase orders, validation costs, and change requests and B) it would be embarrassing to admit such a thing. Nevertheless, that’s the reality. They gambled. I witnessed from nearby and, just like my two Vegas friends, they bet on opposing squares. One wagered with the design artists while the other bet with the usability practitioners. The executives literally decided they needed to win big, sauntered up to the [conference room] table, and slapped down some serious cabbage on red or black. They lost … big.
Why do I consider this gambling? Isn’t it just product development? Isn’t action in our fast-moving world a better risk than inaction?
To understand my beef, you’ll need to connect the dots with me: Companies have a successful product when customers buy it for more than it cost to make it. Right? We the public buy those products to fulfill a need or a longing (or both). The way to know you are heading towards an agreeable ending is to test your product early and often with customers to make sure they need or long for it (well before you’ve invested millions). Seems pretty logical, but you would think it was crazy talk if you lounged in some automotive-engineering halls. The two aforementioned user experiences were never truly tested with customers on reference hardware that approximated how production hardware would function until nine figures had been spent. And, even then, they were tabletop prototypes with only a portion of the functionality, graphics, etc. They presumed the customers would like it, and they kept writing checks.
In Alan Cooper’s famous book The Inmates Are Running The Asylum, he speaks of how some executives in Silicon Valley measure success based upon quantifiable measures such as number of features delivered on time with a “spray and pray” method of getting technological features to market. “Put a little money into a lot of investments and then hope that one of them gets lucky,” he writes, later equating this misguided direction to Stalin’s clearing of a mine field by marking a regiment through it. “Efficient, humanitarian, viable, desirable? No.” Humorously, this book is found within both company libraries, and many development engineers from both camps have been sent to Cooper’s training seminars.
So what happened? How did the bouncing ball not land on their square? I would say there were three reasons:
- A Stitch in Time
- The Gambler’s Fallacy
- Betting on Just One Color
A Stitch in Time
As the saying goes, “A stitch in time saves nine.” Or, in other words, it's more efficient to correct a problem in the beginning, rather than temporarily ignoring it and investing a large amount of effort to fix it later. This philosophy dates back centuries, however the sensibility of this phrase hasn’t permeated as far and wide as it must. How does that apply here? In his book Principles of Software Engineering Management, Tom Gilb presents the cost-benefit ratio for usability testing as $1:$10-$100. “In development, correcting a problem costs 10 times as much as fixing the same problem in design … [which] costs 100 times as much relative to fixing in design.” These automotive companies decided not to spend the figurative $1 upfront on building a modeling prototype of the user interface, testing it with real customers in their native environments, and confirming things were headed down the correct path.
That’s OK if you somehow have designed the perfect cockpit experience without a single problem, which, in all likelihood, was crafted around a nebulous description by an executive … but wait. Another study in Thomas K. Landauer’s The Trouble With Computers: Usefulness, Usability, and Productivity shows that the average UI has some 40 flaws. “When usability is factored in from the beginning … [it] can yield efficiency improvements of over 700%."
It would seem their wager wasn’t a great one.
I literally heard a director of advanced development in one of these companies say, “No. We need to spend our resources on something that’ll get us more revenue rather than prototyping our user experience.” And a manager at the same, misguided company say, “I don’t have time or budget to worry about human factors.” They quickly surveyed the land and saw a down economy, a dismal balance sheet, and a mandate from above stating something competitive must be produced by x-date and immediately marched into that metaphoric mine field. In the midst of that, they hired the “essential” engineers and didn’t consider revamping the process, the toolset, or the mindset. So, yes, they are presently paying $100 on the dollar to fix these issues that were avoidable with a stitch in time.
The Gambler’s Fallacy
The Gambler’s fallacy is when an addict says, “I’m about to get extremely lucky because there’s been a string of bad luck and … well … the Law of Averages says that things will work out.” Of course, the truth is that each event has singular probability and isn’t influenced by previously bad luck. The automotive manager believes this change request will be the last payout and the problems will be solved since, surely, this cannot go on forever.
So, in this Gambler’s fallacy, who wins? The casino, of course! And the casino in this automotive exchange is the supplier. The automotive customer produces a monolithic—yet inadequate—paper specification that has only considered a fraction of the true use cases, the supplier quotes delivering that specification, and then the auto company discovers months later that: A) the 3000-page spec was woefully insufficient, B) the previously selected processor is underpowered, and C) the product’s experience is not aligned with customers’ wants.
So another change request is quoted by that now-opulent supplier who delivered the ill-guided debacle in the first place. Let’s examine three Tier 1 suppliers to the aforementioned roullete addicts: Bosch, Panasonic and Microsoft. Bosch Limited stock is up 30% over the past year with a gross profit of over $19 billion and several key executives making over $35 million per year. Panasonic’s stock hasn’t fared so well, but the company had nearly $24 billion in profit. Microsoft is up nearly 10% over the past year and had a gross profit of over $56 billion with a profit margin over 23%. And although these numbers are staggering, they will continue to rise until the Gambler’s fallacy is abandoned as a philosophy. Should we blame the supplier (or “casino”)? No. In any supply-and-demand scenario, if there were no demands for change requests, then the folly would end.
Betting On One Color
Maybe the craziest part of this whole game is the belief that executives must gamble on just one color: either the design artists or the usability practitioners. Imagine a roulette table where you could bet on both black and red! Seems silly, but in this instance, they could have. Styling and usability need not be diametrically opposed. They can coexist. In their winter/spring issue from 2005, Princeton’s EQuad News said it best:
It is no coincidence that Leonardo da Vinci is both the best-known engineer in European history and one of the best-known painters in the world. But the gulf between art and engineering continuously widens in the popular imagination. The rapid expansion of technology in the last two centuries has led to the widespread perception of artists and engineers as inhabitants of different worlds: one of instinct and passion, the other of pragmatic calculation. While the advances in science have made it more difficult for the artist to continue the Renaissance tradition of extending the artistic imagination to technical pursuits, there are no obstacles to restrain the engineer's pursuit of aesthetic gratification.
Design artists and usability practitioners need not be two separate teams in two separate camps. They can be one team, or one office, or even one person. Pull the artists’ renderings into a modeling tool, test that early and often with users, and iterate together. No one has ever said user-centered design must exclude the graphic design team. I’ve assisted a third competitor and have seen the Promised Land. This company wisely invests on a metaphorical roulette table with only one color and, whallah, has gained market share. It is possible.
The First Step: Admit there is a Problem
Just like that leeching I witnessed on a Vegas roulette table, I hope at my very core for the automotive bloodletting to stop. I watched the Superbowl commercial with Clint Eastwood and, as an American, I truly believe we are at half time and can come out with a winning strategy. Many opined that Obama’s decision to resurrect the automotive industry was ill-advised at the time, but that fateful act has averted what could have been an economic heart attack at the core of this great country.
That said, his assistance has allowed the aforementioned hubris and gambling addiction to sustain. To survive, these great automotive companies must help executives transition from gamblers to investors. The first step toward doing that is admitting that there’s a problem and seeking help.
Roulette image courtesy Shutterstock