Beyond Data (part 1): J. K. Rowling, Shakespeare, and the Sorcerer’s Decision
I rode the train on a beautiful spring day through New England. Arriving at the station, a few guys from our partner company greeted me happily; we were going to win this one. We chatted about the customer on the drive across town: “They’re really turning over a new leaf”…”Very innovative”…”Hungry for new ideas”.
When we arrived, it was a big meeting: folks from several departments, some who hadn’t met each other before. The IT guys carried stacks of paper: the data model…the spreadsheets. As usual, the first part of the meeting was a data dump, syncing up. And, as usually happens, it was mostly about data. Our customer took us through a great database schema on the whiteboard…his colleague handed out reports. Great!
[bctt tweet=” ‘The decision is only as good as the data that supports it.’ It was bound to be said. “]
Because every organization we work with does sooner or later. But it’s not always true.
Here’s what happens: as we get into decision modeling engagements like these, we realize that all data isn’t created equal: some fields matter a lot…some not so much. And often human expertise, where there’s no data, makes all the difference.
The most extreme example of this comes from winner-take-all effects that dominate national economies, markets, and more. I’ve found that system dynamics matter far more than data in many situations, but are systematically overlooked. So a lot of the data that you might spend millions to gather, cleanse, and display is irrelevant to your competitive advantage, especially for big decisions.
[bctt tweet=”Not all data is equally important to your organization.”]
Consider Harry Potter and the Mystery of Inequality, where Alex Tabarrock writes that J.K Rowling is the first author in history to earn a billion dollars, way ahead of Homer, Shakespeare, and Tolkien. Although the average writer’s income hasn’t increased in the last few years, inequity is increasing: the top is pulling away from the median.
[bctt tweet=” The winner-take all pattern goes way beyond books. And it’s getting worse, everywhere.”]
As explained by Investopedia,
“…the prevalence of winner-takes-all markets is expanding as technology lessens the barriers to competition within many fields of commerce. A good example of a winner-takes-all market can be seen in the rise of large multinational firms, such as Wal-Mart. In the past, a wide variety of local stores existed within different geographic regions. Today, however, better transportation, telecommunications and information technology systems have lifted the constraints to competition. Large firms like Wal-Mart are able to effectively manage vast resources in order to gain an advantage over local competitors and capture a large share in almost every market they enter.”
So to understand market dynamics, and to navigate effectively within them, we need more than data. We need to know which decisions-what we might call the “super levers”-will create a “butterfly effect”: swinging us from winning to losing. In the face of these influences, other factors matter very little. So it’s an important quest.
Indeed, our obsession with data—big and small—has distracted us from where the real solution lies: understanding how decisions lead to outcomes, whether it’s in a winner-take-all market or inside a company, where some decisions, in some circumstances, really matter.
And that customer*: I’m happy to say we won the work, and went on to be invited back half a dozen times for follow-on projects. Their giant database: not needed. Our modeling work distilled their highest-value use case down to a single spreadsheet, with one table about a megabyte in size, and a few dozen fields. That’s all that was needed to drive an analysis that saved a bundle.