The 14 Day Rule for Side Project Ideas

Write the idea down. Then do nothing with it for fourteen days. If it still looks good on day fourteen, you may start. That is the whole rule.

The half life of a new idea

Excitement about a new project decays fast. Mine has a half life of about four days. On Tuesday the idea is obviously brilliant and I can see the whole architecture. By Friday I cannot reconstruct why I cared.

Fourteen days is long enough to outlast that decay. It is not a magic number and you could argue for ten or twenty one. What matters is that the window is longer than your enthusiasm, so novelty is filtered out before it costs you a weekend.

It is a filter, not a delay

People hear this as procrastination. It is the opposite. You are not postponing good work, you are refusing to spend evenings on ideas that were never going to survive contact with Tuesday.

My queue is mostly corpses. Ideas I was certain about, which I now cannot explain. Every one of those is a weekend I did not lose. The rule pays for itself on the ideas it kills, not the ones it approves.

Why there must be no bypass

A waiting period with a skip button is not a waiting period. The version of you that wants to skip it is the midnight version, and that person should not have the keys.

This is the same reasoning behind commitment devices generally, the sort described in Thinking, Fast and Slow: you make the decision once, while calm, and then remove your own ability to revisit it while excited.

Running it

• Park every idea the moment it arrives, with a one line pitch. Capture is cheap and stops the idea rattling around.
• Never look at the queue while you are mid project. Browsing it is how you talk yourself into switching.
• On day fourteen, read the pitch cold. If you cannot explain why it matters in one sentence, delete it.

The gate is built into One Project OS, which refuses to promote anything younger than fourteen days and does not provide a way around it. Pair it with a written definition of done and the survivors tend to actually ship.