Eliminating uncertainty: how to really improve your use of business analysis techniques
Could an 18th century statistician improve your business analysis techniques? Ask Nate Silver, the thirtysomething modern day statistician who these days is accorded rock star status for his successful predictions in the world of baseball and politics.
Famously, Silver rose to prominence having left a job as an economist at business advisory firm KPMG in order to apply so-called ‘sabermetric’ business analysis techniques to predict baseball outcomes.
Politics came next, and Silver not only successfully predicted the outcome of 49 out of 50 states in the 2008 United States presidential election, but went on to call every state correctly in the 2012 election.
At which point, you’re probably wishing that your business analysis techniques could be as insightful.
The good news: as Silver is fond of pointing out, there’s little that he does, that other practitioners of business analysis techniques can’t also do.
Which includes ignoring some analytics thinking that’s currently very fashionable—such as machine learning, for instance—and applying basic statistical principles discovered by that same 18th century statistician, Thomas Bayes.
There’s more to business analysis techniques than Big Data
Silver isn’t a big fan of Big Data. More precisely, he isn’t a big fan of the grandiose claims made for Big Data, and of commonly-used Big Data analytics methods.
And while his book The Signal and the Noise—Amazon’s best nonfiction book for 2012, and selected by the Wall Street Journal as one of the ten best books of nonfiction published in 2012—makes a powerful case for collecting large data sets, it deviates sharply from conventional thinking in terms of how those data sets should be utilised.
Which is where 18th century statistician Thomas Bayes comes in. Because Silver is a fan of Bayesian thinking, and specifically, of what statisticians call Bayes’ Theorem.
And in terms of business analysis techniques, the message from Bayes is clear: think probabilistically, and in terms of margins of error and confidence limits, rather than trying to mine data for absolute facts, and absolute certainties.
More data = better informed business analysis techniques
Here isn’t the place to reprise Bayes’ Theorem. In any case, most statistics courses cover it—usually in the material relating to probability. So if you’re interested in business analysis techniques, or have any background in applying business analysis techniques, then you’ll almost certainly have come across it.
But its essence is simply stated. Data provides us with information, and that information can be used to revise our understanding of the probabilities surrounding specific outcomes. In terms of business analysis techniques, the more data that we can gather—experimental or otherwise—the more we can therefore feel our way towards the ‘right’ answer. But it will be an answer expressed as a probability, not an absolute fact.
“A nice attribute of Bayes’ method is that over time you converge toward the correct result,” is how Silver put it at a Gartner Business Intelligence Summit in Dallas. “People can begin with different beliefs, and if they abide by Bayes’ Theorem, in the end they converge towards a consensus as more and more data is accumulated.”
Put another way, business analysis techniques are about reducing uncertainty. Better techniques—and better data—can therefore achieve bigger reductions in uncertainty.
But what remains is a probability. Absolute facts are rare.
Revise, test, refine—but guided by business analysis techniques
So what does this mean for business analysis techniques in practice? And the use of business analysis techniques in your business?
Again, it stems from understanding that absolute facts are rarely encountered, and that business analysis techniques should be used to point us to probabilities.
Probabilities that can then be tested, further eliminating uncertainty.
As Silver told the Gartner conference, it’s this testing—informed of course by business analysis techniques—that takes businesses closer to a true understanding:
“If you look at what Google does, for example, they’re running literally thousands of experiments every year by tweaking their search products and by adding or dropping different products that they run. What they realize is that if you’re in a data rich environment, there’s no substitute for testing your models, your hypothesis, on real customers, on real data.”
To find out more about the tools you need to enhance your business analysis techniques, download our free guide below