Beware of a Single Data Point
We like single data points, and they work well for simple comparison in simple systems. They also work well in games with scoreboards. In business, however, a single data point should signal alarm bells in your head. Your mind should be screaming, “Does not Compute! Warning! Warning! — Danger Will Robinson!”. Okay, so your name isn’t Will Robinson, so replace your name with his.
Before we get into why a single data point is dangerous, let’s first discuss why a single data point is so alluring. A single data point gives our brain a shortcut. It is easy to digest, and it easily fits into our mental models. This shortcut is what Daniel Kahneman calls System 1 thinking or “Fast” thinking. System 2 is our brain’s “Slow” thinking, which we will discuss later in this post. I want to write beautiful prose on the difference between System 1 and System 2 thinking, but it is easier for me to use my System 1 Blogging brain and Google a better explanation. And it worked. I found a great description in the New York Times Book Review of “Thinking Fast and Slow.”
System 1 uses association and metaphor to produce a quick and dirty draft of reality, which System 2 draws on to arrive at explicit beliefs and reasoned choices. System 1 proposes, System 2 disposes.
So System 2 would seem to be the boss, right? In principle, yes. But System 2, in addition to being more deliberate and rational, is also lazy. And it tires easily. (The vogue term for this is “ego depletion.”)
Too often, instead of slowing things down and analyzing them, System 2 is content to accept the easy but unreliable story about the world that System 1 feeds to it.
So there we have it, a single data point “produces a metaphor and a quick and dirty draft of reality.” It is like crack for the business’ soul. But why do we do this? As mentioned above, it is hard to get the “System 2 to” to slow us down and think. And it is even hard fighting the exhaustion of explaining ourselves constantly in meetings and emails. So we all get tired and let our lazy “System 1” drive the bus. For those nerds out there, you might remember that it took 7.5 million years for the computer to calculate the meaning of life in “The Hitchhiker’s Guide to the Galaxy.” The answer was a famous System 1 response 42.
In statistics, you will hear it preached that causation does not equal correlation. We all know this; unfortunately, we miss that our brain doesn’t work this way. Our brain is oppositely and is always looking to turn causation into correlation. The easiest way to see this is to look at people that believe fake news. Fake news only increases the existing belief system. Thank you, System 1.
Before we get high and mighty, let’s do some “System 2” work. First, how often do we only listen to what we already believe? If your response is “I am an open-minded person,” I have two questions.
1) When you don’t like someone, and they share an opinion, how often is your first response “I don’t like that idea.” This thought is a System 1 response
2) When someone you like shares their opinion, your initial thought is probably, “Wow, that is smart.” This praise is also a System 1 response.
Ouch, it hurt writing those two sentences. Guilty as charged.
Back to the task at hand, the evil single data point. What do we do with it when we notice it? The first step is recognizing them in the wild. The next step is finding out a baseline measure. For example, you can become a trillionaire for $5 on eBay Inflation in 2009 in Zimbabwe led to the creation of the worthless 100 Trillion Bill. Last but not least, the final step in killing an evil single data point is finding a corresponding data point(s). Below is a sample list.
Corresponding Data Points
- Revenue has Profit Margin
- Order Fill Rate has Inventory Cost
- Surveys have Response Qty and Rate
- Sales have Churn
- Resolution Time has Request Type
- Email Open Rate has Qualified Leads
- All Datapoints have a historic
As you can see, the second data point adds perspective and prevents data manipulation.
Most data points bubble up from the front lines, and when you can, make sure you include at least two data points. When providing data, do your best to slow down and use your “System 2” thinking. Remember that people are trying to find the signal in the noise at every business level, and the signal gets fainter and fainter unless it comes with a corresponding data point(s). So when you share data, make sure your data tells the true story and isn’t a “dirty draft of reality.” Keep in mind that as your data points are flying around the business, they lose context. The best reports should be easily digestible bytes for your colleague’s “System 1” thinking, and their lazy “System 1” has to interpret this data correctly. So with your data points, take your time and put your “System 2” to work, slow down your thinking, and follow the Navy Seals saying, “Slow is Smooth and Smooth is Fast.”
