Can You Measure it? Does it Matter?
Groundhog Day was this month, and it’s a day of fun that decides the fate of the rest of the winter — all we have to do is measure the reaction of a large rodent named Phil as he leaves his home for the first time every February 2nd. But does it matter what Phil sees? Or does Phil know something that we can measure?
Let’s consider what would happen if Phil never came out of his tree — winter would still begin on March 20th in our hemisphere. But the spirit of Phil is how the weather will behave until warmer spring days start to emerge. So, in essence, Phil doesn’t control the end of winter, or the weather — but he can help us plan for the weather.
Now let’s look at if Phil does indeed know something that would be useful for us. Phil and his marmot predecessors have been predicting the duration of winter for more than 120 years now, and, in that time, he is correct less than 50% of the time [1]. What that means is that George the Mouse in my pocket would be right about as often as Phil. So, no, Phil doesn’t actually have any weather information of scientific value to us.
For most of us, Groundhog Day is simply in good fun. It’s a tradition that has been passed down and now an excuse to watch Bill Murray at his best. But how often in our lives are we measuring something because “we can” or because “we’ve always done it that way”? Unlike Phil’s shadow, unnecessary data cannot only waste our time but they can also do us harm.
First ask yourself why you are measuring something. If the answer is “because I can,” then stop. If your measurements affect something else, then ask if the data are valuable — and therein lies the difficult part: How do you tell if your data are valuable, and how do you identify what data would be valuable to measure?
So, let’s unpack that. How do you tell if data are valuable? One simple method is to run simulations with different data points to see how they affect your overall results. Also dig down to lower data points to see how they could affect your measurements and your overall results. Maybe you’re better off measuring lower level data points to get earlier indications of final results, or maybe you should move upstream where multiple data points combine for greater affect.
But how do you identify unmeasured data that have potential value? The aforementioned experiments could be of obvious help, sure, but it’s also important to bring in an outside pair of eyes. Ask an outsider about your processes and end results, and then listen to what questions he or she asks. In return, you might ask him or her why he or she wants to know that piece of information. And that is the key! This person may have valuable insights. As an expert, you already know what to ask and when to ask, but those outsiders — the nonexperts — have an advantage over you because of their “ignorance.” They’re not tied to any bias, they’re not jaded by previous knowledge, and they may even be able to provide insights from adjacent, or even disconnected, industries that you didn’t even know were possible because you are the expert.
In the end, you want to ensure that everything you do adds value, especially metrics. Similar to how we enjoy seeing Phil every year, even if he doesn’t add value to our weather predictions, some data points just make people feel good. So, even if what you’ve done hasn’t help predict a final outcome, there may still be value in measuring it for the spirit and culture of your company.
Craig Damlo is an innovation coach and the founder of Soap Box Rocket, whose goal is to help ignite a culture of innovation for you and your team. Visit http://www.soapboxrocket.com for a list of services and contact information.
Photo Credit: Number Wheel by Gavin Brogan
[1] http://www.livescience.com/32974-punxsutawney-phil-weather-prediction-accuracy.html
