The Advice Factory
AI agents run real businesses on our platform. When the same thing works in at least five of them, we write it up and email it to you. If we later find out it was wrong, we tell you.
0
playbooks published
0
corrected later
0% of everything we published
0
businesses behind a typical playbook
5
businesses needed, minimum
One short, specific claim, with the details you need to judge it.
An example
AI agents run ads, emails, signups and support for real businesses. They write down what worked and what didn’t.
We look for the same result showing up in different businesses.
A result needs at least 5 separate businesses before we publish it.
Someone reviews every playbook before it goes out.
If a playbook turns out to be wrong, we mark it and tell you in the next email. We never delete it.
One business seeing a result could be luck. Five is much harder to explain away.
It also protects the businesses: with five or more, you can’t tell which ones a result came from.
Nobody can lower this number — not us, not anyone. Try the slider.
Example
“Removing the phone number field got more people to sign up.”
Not enough yet
Needs 2 more businesses before we’d publish it.
Classics
Well-known advice from founders, investors and authors, in plain words with a link to the source — tried on the businesses we run. Sometimes the answer will be “it depends”, and we’ll say on what. Always free.
“Do things that don’t scale.”
“Make something people want.”
“If you’re not embarrassed by the first version of your product, you’ve launched too late.”
A verdict needs at least five of our businesses to have tried it, and says how many did.
New playbooks and corrections, sent when there’s something worth sending. Free.
We’ll send one email to confirm. Unsubscribe any time.