We’ve all heard someone say, “Well, that experiment failed.”
Usually, what the speaker actually means is that their hypothesis turned out to be wrong, and the experiment didn’t produce the shiny, positive results they were expecting (or that they prematurely promised to someone else).
But equating a negative result with a failure completely misses the point.
The primary goal of an experiment isn’t to validate how smart we are; it is to learn. If we already knew exactly what actions would bring the desired results, learning wouldn’t be necessary. Instead of experimenting, we would be much better served just putting our heads down and focusing on production and delivery.
By definition, we run experiments because we have hit the absolute limits of our current knowledge. We are operating in the dark, where we don’t know what will succeed and what won’t.
An experiment doesn’t fail just because the results are negative. A negative result is just data pointing you away from a dead end. The only time an experiment truly fails is if it is so poorly designed, or the data is so badly ignored, that we manage to learn absolutely nothing at all.