As Jim Goodnight patiently sat through a press interview, flying from North Carolina to Las Vegas on his private jet, he leant forward and made a small admission. Sometimes, admitted the billionaire mathematician and founder of global software giant, SAS Institute, his heart overrules his head and, mid-hand, he abandons the system he has painstakingly worked out for beating the odds at three card brag.
To be fair, it is not exactly a headline-making confession. But to this interviewer, it was a small victory, a brief insight into a bigger debate. It might even be called a ‘blink’ moment. But first, the background.
SAS, Goodnight’s company, is a big proponent of analytics – not just any analytics, but what we might call ‘hard analytics’ – the application of software to the right data, in real or near real time, to strongly inform, or even dictate, daily decision making. This is a fair bit more advanced than a lot of traditional business intelligence, which is more about what used to be called ‘decision support’. People might run a quick analysis, or look at some charts and break-outs before going ahead, but it is they, and not the machines, that make the decisions.
Hard analytics puts real faith in the data, and it is very much in fashion (and will be featured in greater depth in the January issue of Information Age). Some pioneers, among them Amazon, Tesco, Barclays and Marriott Hotels, are cited by analysts as companies that are “data driven” or that “navigate by numbers”, barely making a single misstep as they progress through their turbulent markets.
At these companies, or perhaps we should say, at the next-generation companies that are modelled on these, when the computer says “no”, it really means no. Executive gut feeling, hunches or bloody minded determination to succeed are not enough to swing a decision. Decisions are made on evidence, and the results are there to prove it works.
The analytics driven turnaround at Harrah’s Casinos is a case in point. Helped by the theatricality of its CEO Gary Loveman, Harrah’s has become the poster child for the successful use of predictive business intelligence. Move over Wal-Mart Stores with your beer and diapers.
Loveman at Harrah’s argues that too often in business, decisions are made by managers without the right information: “When you go to a doctor, he doesn’t guess. He’ll get some data. He’ll run some tests”. In his casino, he says, there are three ways to get fired: steal, get arrested – or not use a control group.
At this point, some readers might be beginning to feel uncomfortable. Will the computers really be given control? What about gut feel or experience? Gary Cokins, a top consultant at SAS and author of several books on performance management, is unequivocal. Managers who think like that are dinosaurs, he says, and the sooner they are gone, the better.
Tom Davenport (management guru and author of Competing on Analytics: the New Science of Winning) reiterates the sentiment by citing Ed Deming, the quality guru: “In God we trust; all others bring data”. Yet another analyst explains what this means: business people need “to make the transition from epiphany to evidence”. In other words, stop using gut feel, and start doing the analysis.
The evidence in favour of this approach is undeniably strong, but as a sceptical journalist and attendee of many board meetings over the years, it is difficult not to wonder how much this is really followed, even by those who believe in it.
And what about the more sceptical executives? Who in business has not seen good evidence ignored, twisted or overridden, or seen decisions made without good evidence altogether? It is, after all, asking executives to put an enormous amount of trust in systems and machines that have been designed and programmed by mere mortals, and using datasets that, in the past, have so often proved to be flawed.
Even SAS founder Jim Goodnight admits to this, albeit only in the context of a card game. Nevertheless, he is no ordinary dabbler; he has been honing his statistically based system for beating the dealer for several decades.
Executives who are worried that computers are about to take control can take refuge in one of the surprise best-sellers of 2006, Malcolm Gladwell’s Blink. Gladwell, who wrote The Tipping Point, finds that the human brain is capable of making accurate “gut-feel” decisions based on the thinnest of evidence.
But there is a drawback: It’s not always possible to tell when the gut-feel decision is right or if an error has been made – it’s all about trust. But then, who really trusts the data in their computers?