Successful businesses run experiments. Constantly. Harvard Business School professor Stefan Thomke found that companies like Amazon, Google, and Booking.com run thousands of experiments per year. They don't guess—they test. And they grow faster because of it.

Scott's not Amazon. But yesterday, his team ran an experiment—and he's sharing it.

The problem: Churn. One of Scott's SaaS businesses tracks churn monthly. The report shows the number, but not the why.

Churn applies to everyone: Training companies, land investors, service businesses. Anytime you have to go get new customers, you have churn.

The excuse: Someone on the team quoted Alex Hormozi—"small businesses go out of business." Scott's response: "You can't do that."

The experiment: Scott gave AI access to the data. The insight: customers with reviews cancel 58% less. One package didn't allow reviews—and had the highest churn. They turned on reviews for that package yesterday.

The hypothesis: Churn will drop. They're measuring over 30, 60, 90 days. Scott will report back.

The Slack example: Slack found that the more colleagues you added early, the more likely you were to become a paying customer. That's why they push "bring your team."

The lesson: What can you do early to help your customer succeed? When your customer succeeds, you succeed.

The challenge: What experiment could you run this week? Change one thing. Measure before and after. True or false—did it work?

Got a business question? Ask Scott here: scotttodd.net/ask