That Randy Brown

Fidelity Thresholds

When introducing a new process, in order for the change to be successful, I have a theory that the following cases must be true:

  1. The process must be too easy to avoid (or not too easy to avoid, depending on how you read that clause) or get wrong. I have a friend who is a brilliant engineer who calls this 'the pit of success', a phrase I quite like. In this case, the process is probably best represented by one step (not one additional step) that can be done at any time and still be successful. An example is 'drink water'. This is too easy to get wrong. You can't drink too little and there's no timeline associated. If you do it at all, you are successful. It is not hard not to drink water, but it's also not hard to find drinkable water. Most of the people reading this probably have 5 faucets within 2 minutes of them, all of which most likely provide drinkable water. This would only be easier if it were harder to avoid drinking water.

  2. It must be successful at low thresholds of fidelity. If I follow the process, it should provide value in a low threshold of participation. This increases the likelihood of future engagement. At worst, it should introduce no friction at any engagement. Drinking water, for instance, helps us move more, if only for the day we drink, and can help with energy levels. In an ideal world, a person would notice the benefits of drinking a glass of water infrequently, and would then drink water more frequently. For a process to be propelled to habit or routine, it must yield results on every action, even if those actions are infrequent. This is critical to early sustainment.

  3. Related, a person's capability for delayed effects is directly tied to their sustainment of other processes. Having been able to pay off one thing over time is an indicator in being able to pay off another thing over time. This is the core of our credit score system (which is worth a whole other note). In the previous example of drinking water, the benefits of water can be timed to an hour or so, which is likely too long to yield good feedback loops. Under 5 minute effects might be better for reinforcement, but those are likely to be too subtle to recognize. As we become more comfortable with current processes, new processes can introduce slightly longer effect waves. Stretching before bed or waking up is likely not to yield an immediate shift in physicality, nor is reading ten pages one day likely to produce new thinking patterns. To sustain those, first, we probably need to drink water or get a good night's sleep.