My best experiments do this one thing...

Most experiments fail, but my winners have one surprising thing in common…

As I was writing my book, my editor noticed something weird:

I included examples of the best hypotheses I’d seen, ones that led to the biggest breakthroughs.

“Did you notice these are all negative?” she observed. 

For example: “We believe people are not signing up because…” or “We believe people are not activating because…” etc.

“Is there any reason these are all phrased in the negative?”

I hadn’t noticed, but once she said it, I couldn’t un-see it.

I tried to find a strong “positive” hypothesis. I couldn’t.

Here’s why:

Growth experiments always try to get people to start doing something – click, sign up, buy, upgrade.

And the best experiments are based on a theory about why they’re not doing it.

For example: 

• “We believe people aren’t starting free trials because they’re afraid they’ll forget to cancel and will be billed. Fix: Add 'no card required' next to the free trial button.”

• “We believe developers aren’t clicking ‘Book a demo’ because they hate talking to salespeople. Or anyone, for that matter. Fix: Give SDK + test API Key with no email required.”

Users don't need more reasons to act. They need fewer reasons not to.

Simple next step

1. Look at the experiments you’re planning

2. What action do you want people to take?

3. Why aren’t they doing it already?

4. Find the blocker. Address it.

Stop asking “what would convince them?” Ask “what’s stopping them?”

I hope this helps!