37% exploration. 63% execution. It's not a productivity hack. It's optimal stopping theory applied to how you spend your time. Most people have the ratio inverted.
The optimal stopping problem
In mathematics, the optimal stopping problem asks: when should you stop searching and commit? The answer, derived from the secretary problem and other variants, is roughly 37%. Spend 37% of your available time exploring options, then commit to the best candidate you've found so far.
Apply this to productivity: spend about a third of your available energy capturing, exploring, and planning. Spend the other two-thirds doing.
Most knowledge workers have it backwards
Taking notes feels productive. Reading and researching feels productive. Setting up a new system feels productive. None of these are actually output. They're preparation for output — useful to a point, then a form of sophisticated procrastination.
The symptom: a perfectly organized note system and a set of tasks you never close.
Tasks are RAM. Notes are ROM.
We think about this as two modes. Notes are ROM — durable knowledge, slow to write, slow to change, high value when referenced. Tasks are RAM — active working memory, fast to create, meant to be closed.
RAM runs out if you never flush it. A task list that only grows is RAM with no garbage collector. The discipline is closing tasks, not just capturing them.
The manager and the employee
Good managers plan well and then step back. Good employees execute without second-guessing. You need to be both. The mistake is being a bad manager — over-planning, over-organizing, changing priorities every day — which makes the employee version of you impossible to trust with a sustained project.
Set direction once. Execute. Adjust at natural breakpoints, not constantly.
How Owldo is built around this
The inbox handles the exploration side: capture fast, no decisions required, start messy. The task side handles execution: clear list, schedule view, Pomodoro timer for focused sessions, Eisenhower matrix for prioritization.
The app is designed so the exploration and execution modes are separate but connected. Not two apps, not one app with no distinction. The line between thinking and doing should be visible.