I have read dozens of books on strategy, experimentation, and the role of failure in teaching us. I have tried a lot of what I’ve read. What it boils down to is this:

1. You have to have a broad sense of the direction you are going in. Various books call this various things – intent, vision, whatever. The end-goal may or may not be crystal clear – in most cases its quite fuzzy. But you have to have some sense of this to orient yourself.
2. You have to have some action. You have to get in the car and drive in the direction you want to go. Sitting still until the path is crystal clear is often a recipe for getting nothing done at all. We all want the Google Map into the future but usually don’t have it. We have to do things the old ‘Go west, young man’ pioneering way.
3. You need to have a frequent way of testing experiments. What doesn’t work should be stopped and changed. This is analogous to making frequent compass sightings when you are orienting yourself across a path. You take a sighting, walk toward a landmark, and take another sighting. Likewise, in work, you make an experiment, learn from it, and either iterate or try something different. None of this is rocket science. The thing I’ve found most helpful is increasing the frequency of experiments. If you can run a daily experiment, you are 365 chances to get somewhere in a year–vs a weekly (52 chances) or monthly (12 chances). But, in order for an experiment to teach you anything, you must: (1) make failure survivable
(2) define the experiment such that you know if it’s a ‘win’ or a ‘fail’ (or, in my own work, a ‘knock it out of the park’)
(3) define the experiment in such a way that you know what you are trying to learn from it. Another thing I’ve found helpful: have a list of rules or heuristics you use when making initial experimental decisions. I’ve been collecting a personal list of 100 things to remember when thinking. —