It is 1943 and you are Head of the United States Eighth Air Force. Your mission is to destroy Germany's ability to wage war and free up the skies for the Allied Forces. The problem is that your Bombers are being zapped out of the sky like mosquitos drawn into the fire in a summery night.
German anti-aircraft fire and Luftwaffe - the German Air Force - are making your life hard. You cannot afford losing more Bombers and more personnel else the war is lost.
The solution to most problems is data. Data is also the culprit of most problems but that is a different post.
During the WWII, the United States after suffering terrible losses turn to Statistical Research Group (SRG) that had the best statisticians in the world.
The Bombers had to be reinforced with metal in order to cover any weak points. Sadly, you cannot cover the whole plane with additional metal as metal is heavy. The weakest points had to be identified and only these points had to be patched.
The data collection started with analyzing the location of the bullet holes that were found on the Bombers that made it back safely. The obvious idea was to patch the areas that had the highest concentration of bullet holes.
More bullet holes around the wings? Patch that part. More bullet holes around the tail? Patch the tail. Data driven decisions at its best.
There was one critical issue with that theory: The US collected data from the planes that made it back, the planes that survived and obviously weren't blown out of the sky. If anything, they should have been looking to patch the areas that had no bullet holes at all. The planes that went down were never included in the data.
Abraham Wald from SRG suggested, the bombers were going down because they were shot at the engines - an area that had zero bullet holes on the returning planes.
Making assumptions based on a winning, surviving dataset is called Survivorship Bias and it can be devastating for Entrepreneurs.
It is easy to idolize the person that worked for 5 years in a bad idea but at the end made it and now is a multimillionaire. Or the Entrepreneur that left a well paying job and decided to build a new company and now he has more money than the budget of a small European city.
As Entrepreneurs, we learn to listen to the data but what happens when our whole dataset is made up from "planes that survived"? What happens when only 1% of bombers make it back?
We all have blind spots. Every single one of us. We become better by making these spots as small as we can. We do that by stepping back and looking at the big picture instead of getting closer and closer.
From up close, even a dot seems like a huge black hole.
The first step in finding what "brings us down" is start looking for what might be missing. A good example is customer reviews. Relying only on customer reviews is a terrible idea to improve your business. I know a Startup that had brilliant reviews but was bleeding customers. The majority of these customers never cared to provide any feedback - they just hated the product.
The same happens when you ask for opinions from your friends or family. Even if they are hard on you, it is no comparison with the feedback you will get from strangers that don't care about your feelings. I am not saying that "all feedback is good", you still have to filter it but at least collect as much as you want.
Unless you like getting blown out of the sky.
- Thu 03 December 2020
Get weekly tips in your inbox.
Whenever I build a new product, I follow this advice:
"Make the simplest version of your product first. Learn and iterate fast."
I am building a course for Founders and I am coding my own platform for it (it's a long story) and the simplest way to see if the …
No boring, ordinary stories.