Data-driven approaches to achieving great things have been in vogue ever since Moneyball told the story of how the unfashionable Oakland Athletics baseball team were transformed by manager Billy Beane's data driven approach to recruiting and overseeing the team.
It's an approach now widely adopted in other sports, not least of which is motor racing. In his latest book Big Data in Practice, Bernard Marr tells the story of how data powers the Lotus Formula 1 team.
The team are collecting petabytes worth of data to allow near real-time adjustments of the vehicle during testing together with extensive use of simulations to overcome the limits on driver testing imposed by the governing body of the sport.
Big Data In The Workplace
Whilst sport has taken to using big data like a duck to water, it has taken longer to establish a toe hold in the workplace. At least that's the impression anyway, but a recent report from the Institute for Corporate Productivity (i4cp) suggests those assumptions might need to be re-calibrated.
The study, called The Promising State of Human Capital Analytics, reveals that nearly 70% of executive teams are using people-based data to drive their businesses in some way.
"Successful companies tend to be those that purposefully use data to anticipate and prepare rather than to react to daily problems," the authors say. "The future focus of professionals in the human capital analytics field will increasingly be on using analytics to guide strategic decisions and affect organizational performance."
This attempt to better see the future is typified by a platform I wrote about earlier this year. The site, called PredictHQ, monitors both global and local events to try and give managers better insight into things that might influence their business.
Gaining Management Buy-In
Despite the apparently strong levels of usage however, there seems to be a bit of work required in convincing executives of its use. The i4cp paper reveals that 43% of data scientists believed they had the support of their senior management team.
A study published last year provides a glimpse into why that might be. When the researchers examined the investment decisions of managers at over 1,000 companies, they found that gut instinct tended to drive decisions rather than data.
“CEOs systematically put too much money in projects with high potential upside”, the authors say. “Many CEOs themselves say that ‘gut feeling’ is important for their investment decisions. The problem is, that there is by now overwhelming evidence from psychology and economics suggesting that intuitive reasoning in financial matters frequently leads to biased and therefore suboptimal decisions. Our paper shows that investment decisions biased towards long shots may indeed be a serious problem in many firms.”
Another paper, this time from researchers at Wharton underlines the challenges involved. They examined the use of performance analytics at an American furniture company to allow employees to compare their efforts alongside their peers.
Whilst most employees believed this comparison would motivate them to new heights, in reality the opposite was the case. The authors suggest that this is largely because when we're ignorant of our standing, we focus more on the task at hand than on our ranking, whilst also operating as a team rather than individuals.
Data Driven Gains
That isn't to say, of course, that data has no place in the workplace. As with most new technologies workplace analytics is an approach that is still very much finding it's way, and it will be inevitable that there will be mistakes made as we find approaches that work.
The i4cp paper outlines a number of companies that are using data effectively, including HSBC, Intel and Google, together with insights into how they got started.
So whilst it hasn't been the panacea that early advocates perhaps promised us, it is nonetheless something that is unlikely to go away and as we learn to harness data more effectively is sure to become an effective tool in the managers arsenal.
Central to achieving this is broadening the pool of people that are capable of tapping into the data they have available. There is a widely publicized and on-going shortage of skilled data scientists, so DARPA have recently developed a platform to help bridge that skills gap.
Data-Driven Discovery of Models (D3M) helps relative novices by automating much of the complex data grunt work, thus allowing them to create complex models of their own without requiring expert help. Such a platform won't work magic on it's own, but it's another example of how data is being democratized in our workplaces.
I'd love to hear your own experiences about data. Does your own workplace utilize data to operate more effectively? Let me know in the comments below.
