The trouble with data
The data is in, and data has a problem.
Just like the McRib, fads come and go. Hair metal bands. The keto diet. Planking. Knowing the names of your neighbors. While many thought that these movements might never end, sadly they are all relics of the past.
Business and corporations are no different from the rest of culture. They have their own fads: team bonding; the open layout office; six sigma (although I could see this come back in style as Gen Alpha enters the workforce - see their definition of “sigma”); employee equity; triple bottom line. I’ve been around long enough now to spot a fad….with one exception: data.
Data as a word or concept has been around for a very long time and predates my entry into the workforce. Companies have been collecting and using data in one form or another since the town blacksmith and Ye Olde Mutton Shoppe started doing business sometime in the dark ages. But “data” as it is used today, really came into the lexicon sometime in the early 2000’s (I haven’t conducted a formal historical survey, but this just feels right). It seemed to coincide with the rise of software databases and the widespread use of the internet. The former is a way to store massive amounts of information, and the latter is the source for those massive amounts of information that needs to be collected. The general idea was that by examining “big data” that companies could make better and more profitable business decisions.
As business began to reap the benefits of these new tools, data, the fad, turned into data the religion. Data seems to be firmly entrenched into business leaders and is taught by the most elite MBA programs (and thus ensuring that we have a deep pool of leaders who all think alike…but that is a different story for a different day).
If data is a religion, then perhaps it is time for its own Reformation. (My grandmother always wanted me to be a Presbyterian minister. While this may not come true, perhaps being the Martin Luther for the Data Reformation can satisfy her dream. So Grandma, this one is for you.)
When Martin Luther posted his 95 Theses on the door of Castle Church in Wittenberg, Germany, he had no intention of leaving the Catholic Church. Much like Luther, I don’t want to give the impression that I find no value in data. There is quite a bit of value in data. But, as with most things, it has its limitations and is not the divine source of truth that so many have come to believe.
Let’s get into it.
To understand the limitations of data, there needs to be an examination of what value data can provide. The first thing is the ability to see and analyze results. By examining the key elements of a project, business managers can get a detailed understanding of what worked, and most importantly, what did not work. It can also provide insight into whether there was an issue with a particular tactic or activity, or perhaps the wrong strategy was followed from the beginning.
For example, I was once brought into an organization to help them build a revenue machine. Six months into the job, I had already tripled my annual target for contacts and leads (MQL’s). Despite this success at the top of the funnel, there were still no sales. By setting up metrics at each key step prior to execution, I could go back and pinpoint the exact problem. An analysis of the funnel revealed that the problem was more a sales issue than a marketing issue. This analysis helped me pivot from a marketing strategy to a sales-enablement strategy.
The second key value in data is the ability to make future predictions. The best way to predict the future is to examine what has happened in the past. With enough data and analysis, organizations can spot trends and build recipes for future success. Data can provide a roadmap for where to invest and double down.
But, data’s biggest value is also its achilles heel. Data can provide an excellent formula for success, but it can only do so if all things remain constant. And you may have noticed, things change pretty quickly.
Let’s take a look at perhaps the most famous example of data being used to create a winning formula: Moneyball. If you are one of the six people on the planet who do not know what Moneyball is, let me give you a quick overview. The small market baseball team, the Oakland A’s, put together a winning roster despite having one of the smallest payrolls. They consistently were contenders for the division crown using no-name players or players who were said to be past their prime. By understanding what exactly led to winning games and examining the data, Billy Beane, the general manager of the Oakland A’s, was able to change the way that baseball was played.
For all their success and division titles during the time they were using “Moneyball,” they never made it to the World Series and rarely won any postseason series (Note: The A’s last made it to the World Series in 1989 which was several years before Billy Beane and his Moneyball tactics.) So with so much success during the season, why could they not replicate it during the playoffs? Fans in Oakland will tell you that the team is cursed. Billy Beane will tell you that his job is to get the team to the postseason and that the rest is “luck.” But the truth is that Moneyball simply did not work in the postseason because baseball was played differently than it was during the regular season.
Baseball has a very, very, very long season. Each team plays 162 from April through September. While teams want to win each time they take the field, going undefeated is not realistic. But in the playoffs, where it’s “win or go home,” the game is played and managed very differently. Managers pull pitchers sooner. Pinch hitters and runners are used more liberally to generate runs. And players simply care more. They know that their bonus is tied to winning playoff games. The A’s used data to be successful in the regular season, but that success quickly changed once the cold winds of October blew them over.
A few years ago I wrote about The Other Side of Midnight and its relationship to Star Wars (see the full essay here). In a nutshell, Hollywood exec’s thought that the big summer blockbuster of 1977 was going to be The Other Side of Midnight. The film had all the ingredients for success: Script from a best-selling book. All-star cast including Susan Sarandon. And most importantly, it looked like other blockbusters from years past. Star Wars was not expected to be a success mostly because it did not fit the formula that Hollywood knew so well. The data simply didn’t support a film like Star Wars making money. We all know how that turned out.
The point is that data, while great at understanding what happened in the past, is not great at looking forward. In a vacuum this may not be the case, but business and people change too rapidly. But that doesn’t mean that there isn’t another method to look ahead, but that is for a different essay.
