Skip to main content

Running Experiments

I think a good thing to borrow from the product world for our lives is the hypothesis/experiment cycle.

It has many names:

  • PDCA (Plan-Do-Check-Act), the classic Deming Cycle;
  • Hypothesis-Driven Development;
  • Experimentation Loop.

But the core process is the same: form a hypothesis, run an experiment, analyze the results, then either adopt change, discard it, or iterate with newly acquired data. Optional step is sharing your findings.

The Ness Labs blog and the Tiny Experiments book present the PACT framework for this: Purposeful, Actionable, Continuous, Trackable. Their formula for experiments is simple: “I will [action] for [duration].”

Here’s an example – and maybe my next experiment: uninterrupted time in the middle of working day. My experiment: “I will block 90-120 minutes every afternoon (time may vary because of some meetings) with Focus mode fully enabled for a couple of weeks”.

It’s good to keep a journal here, separate from a regular one if you have it. Yes, these entries feel like overhead on a daily basis, but they capture useful context for analysis like mood changes, failures, exceptions you’ve made, etc. Experiment could be technically successful even with days you really hated it. That seems like a place to change something.

The most important thing: these experiments shouldn’t become a full-time job. Want to change something? Think about it, design an experiment, run it. But don’t fall into the full-time self-improvement trap, which leads only to burnout, not actual improvement.