The science of winning

In the 17th century, intellectuals across Europe challenged centuries of establishment thinking, and began to question everything that was not supported by facts, proof, evidence and irreducible logic. It was called the Age of Enlightenment, and the new scientific methods triggered huge advances and changes.

It is a grand comparison, perhaps too grand, but there are signs of a step change in modern business, in which a lot of irrational or ill-founded decision-making may – just may – be eliminated.  Some business leaders,  inspired by some stand-out success stories, and armed with huge volumes of electronic data and a rapidly advancing array of new performance management and analytics tools, are beginning to manage their organisations with a new, fact-based rigour. And in doing so, they are not only achieving some extraordinary results, but inspiring a wave of others to follow.

Not for the first time in the intertwined history of business and IT, one thinker, with a book and a Harvard Business Review article, is playing the role of pilot and catalyst. In this case, it is Tom Davenport, a distinguished professor of Information Technology and Management at Babson College in Massachusetts. He sees a “planetary realignment”, with executives seeking to build a culture and a platform to exploit all the information they now collect.

Thornton May, executive director of the US IT Leadership Academy, concurs: “People really are changing how they think. Executive behaviour used to be based on guesses. We’re not replacing intuition, but we’re reducing the decisions we get wrong. We’re making better guesses. We’re eliminating stupidity.”  If every company followed suit, he says, there would be a lot less waste.

Davenport’s book, Competing on Analytics: The new science of winning identifies an elite group of companies that have taken their use of analytics way beyond what has been possible with business intelligence hitherto. They include Harrah’s Entertainment (see case study), Capital One, Amazon, Dell and Wal-Mart. Speaking recently, Davenport also included the UK’s Tesco and Barclays among the global elite who compete on analytics.

These companies all share some key characteristics, including the rigorous use of sophisticated information systems, a culture of fact-based decision making, the use of “a lot” of people with “the very best” analytical skills, a voracious appetite for the collection, analysis and distribution of information, and a ‘test and learn’ culture based on numerous small experiments. Amazon, for example, uses control groups to test the commercial effectiveness of every new service and major web page change.

Such companies also take an enterprise approach to analytics, with centralised groups that work with common technology and tools, under the control of one leader. These groups use sophisticated techniques, manage the data to ensure quality and consistency, and provide analytical services to support the rest of the business.

Tesco, for example, not only has a “digital insight team” that spends most of its time analysing consumer behaviour on its websites, but has also created a joint venture, Dunhumby, that analyses customer and store card owner behaviour. “Everything they [Tesco] do is driven by data,” says David Pool, whose company supplies analytics tools to the retailer.

A wider wave

Not all these elite companies enjoy, and certainly will not forever enjoy, sustained competitive advantage. But it is no co-incidence that they are mostly in robust health, outperforming competitors.

They are only part of a wider change in business behaviour generally. “Tom [Davenport] has focused on the elite companies – the wow factor. But using information as a means to compete is something I see a move towards all over the world, in every vertical,” says Frank Buytendijk, VP of corporate strategy with business performance management software supplier Hyperion. “This move is powering the elite, but it hasn’t properly hit the mainstream. It’s all about being ‘smart’, about combining all [an organisation’s] expertise about how to get from ‘good to great’,” he adds.

There is clear evidence of this wider wave. In 2006, for the first time, for example, business intelligence topped Gartner’s annual list of top ten CIO concerns – pushing security, integration, outsourcing and others down the list. AMR Research, another analyst group, said sales of business intelligence and business performance management software grew to an unprecedented $23 billion in 2006.

Meanwhile, organisations seeking to recruit analytics specialists report a shortage of supply in almost every area – whether it is in quantitative analysis and statistics, skills with the IT tools, or specialist skills in business performance, customer relationship analysis, supply chain or web analytics.

Although there are several drivers behind the analytics wave (several experts cite a change in emphasis towards quantitative skills on MBA courses), new technologies are playing a key role.

Stephen Brobst, the chief technology officer of analytics systems supplier Teradata, says one of these is the adoption of operational, as opposed to strategic, business intelligence – drawing the intelligence from operational systems as events occur. This may mean providing information to decision makers wherever they are, whenever they need it, in real time; or it might involve integrating the information with the applications and processes, taking the human element out of the equation altogether.

“About a quarter of Teradata’s 1,000 blue-chip customers have some kind of operational intelligence capability. The rest are maybe thinking about it,” says Brobst.

Another new and related part of this is the use of predictive analytics. This involves using the business intelligence to predict what could happen and so intervene before an unwelcome outcome occurs.

This enthusiasm for BI is not without problems. According to recent research from SAS, the analytical tools supplier, most investments in performance management and BI do not provide a clear return on investment. Among the reasons why: cultural resistance, poor inter-departmental collaboration, a lack of accountability, suspect data, too many tools and a lack of integration, and even an overload of reports.

But problems implementing analytics should be properly tackled and analytics embraced, Davenport believes. His findings are clear: “Use of analytics is good; competing on analytics is better. There is a strong correlation between being analytically oriented and being successful”.

Kirk versus Spock

Jack Welch, widely revered as the greatest business leader of the 20th century, titled his autobiography From the Gut, intending to convey the value he put on courage, commitment and a willingness to follow things through.

Those values are as important as ever in business leaders. But ‘gut feel’ – an instinctive sense of what to do in the smoke of battle – is no longer the unquestioned virtue that it once was. Indeed, at companies such as Harrah’s, Amazon, Marriott Hotels, Tesco and Barclays, ‘gut feel’, hunches and intuition are frowned on. Barry Beracha, recently retired as CEO of Sara Lee, the bakery group, summed this up with a sign on his desk quoting quality guru Ed Deming: “In God we trust, all others bring data”.

This faith in computers, analytics and evidence is no hunch, says Stephen Brobst, the chief technology officer of NCR Teradata, the data warehousing systems supplier, speaking for many in the modern analytics industry. As long as they are working within their realm of experience, and the problems are fairly defined, “it has been consistently shown that machines can outperform humans”.

Ignoring the data and the computer’s ‘advice’ is, in some companies, a heretical act which could be difficult to justify – especially if regulators, shareholders or lawyers ask for evidence to back up controversial decisions.

Many managers, of course, think that gut feel is the seat of creativity and competitive edge. They see a role for facts and analysis, but only to a point. In the ‘instinct v analysis’ debate, they think that the occasional impetuous Capt Kirk makes a better leader than the reliably rational Spock.

Two recent books have fuelled the debate. Malcolm Gladwell’s Blink, the successor to the enormously influential Tipping Point, describes how experienced people are able to make extraordinarily accurate and perceptive assessments of complex situations with only a very ‘thin slice’ of data. Often, they do this without even knowing how – they use gut feel.

Another book, Money Ball, by Michael Lewis, describes how so called “sabermetrics” helped the Boston Red Sox baseball team win three of four World Series. US information guru Tom Davenport even cites UK premiership football team Bolton Wanderers as a team helped by analytics.

If analytics has invaded sport – an area where ‘guts’ could scarcely be considered more important – then why can it not enter into that most subjective area of all, the arts? Platinum Blue Music Intelligence, based in Barcelona, has a database of every Top 40 hit in the UK and every Billboard Hot 100 in the US since the 1960s. Using what it calls “spectral deconvolution”, it has deconstructed the key characteristics of any song and clearly identified those that have what it takes to be a hit. The technology is reportedly being trialled by some record companies, but has inevitably met with hostility. Guy Chambers, who co-wrote the Robbie Williams hit Angels, told the Guardian: “It’s evil”.

There is, of course, room for some compromise in the ‘analysis v instinct’ debate. Most managers use as much evidence and support as they can from computers and elsewhere, but then make their own decision using their experience and gut feel. But managers who don’t use data at all, or who irrationally override or twist the findings, are “dinosaurs” destined for extinction, says Gary Cokins, a business performance management specialist with business intelligence software supplier SAS.

“For analytics-minded executives,” says information management guru Tom Davenport, “the challenge boils down to knowing when to run with the numbers and when to run with their guts.”

Further reading in Information Age

Part II: Nothing left to chance – January 2007

Science for winners – December 2006

Customers in the tank – August 2006More articles can be found in the Business Intelligence Briefing Room

Pete Swabey

Pete Swabey

Pete was Editor of Information Age and head of technology research for Vitesse Media plc from 2005 to 2013, before moving on to be Senior Editor and then Editorial Director at The Economist Intelligence...

Related Topics