What is a kill rule for a side project?

A kill rule is a sentence you write before launch that says when you'll stop: a metric, a number, a date and an action. "If fewer than 3 customers are paying at day 30, I close it." Written while you're hopeful, it lets you stop calmly later, instead of letting a product fade for years.

Why it has to be written before launch

On launch day you're at your most optimistic, and that's the point. The number you write then is generous. Every day after, it gets easier to move the line, to tell yourself it's early, that next month will be different.

I kept a product alive for 3 years after I knew it wouldn't grow. A booking-reminder tool for driving instructors, with a handful of customers and a backlog of 40 small improvements I added to every Sunday. I didn't keep it for the money. I kept it because closing it meant telling people it hadn't worked. (The customers would have been fine.)

The 4 parts

• A metric. One you'll actually see: paying customers, retention at day 90, the weekly look you skipped.
• A number. Specific. "Some traction" isn't a number.
• A date. Day 30, day 60, day 90.
• An action. Close it, hand it over, or change one named thing and set a new date once.

The rule that killed 2 working products

My most useful kill rule is about me: if I skip a product's weekly slot 3 weeks running, it has to win my attention back or be wound down.

It has fired twice, both times on products that were working. One was a fees tracker for small language schools that met every number I set for it, which I'd simply stopped caring about. I'm not certain the rule was right about both. I'm certain it was right to make me decide.

What killing buys

A slot back, mostly. A dead product kept alive takes an evening a week, a background worry, and the excuse not to start something better. Close it properly, with notice, a refund and an export, and you get all 3 back.

Ship Alone has the full chapter on kill rules. Nobody Bought It has the 2 products the attention rule closed.