I think we've been measuring lead quality the wrong way.

Most startups I talk to still score leads based on things like:

Company size
Job title
Traffic source
Demo requests

But I've noticed those signals often don't predict who actually becomes a customer.

What seems more interesting is behavior.

Someone who:

comes back three times,
explores pricing after using the product,
invites a teammate,
or keeps returning to solve the same problem...

...often looks more promising than someone who fills out a demo form once and disappears.

We're building Autonomy around this idea—trying to identify buying intent from product behavior instead of static lead data.

Curious how others approach this:

What's one user behavior that turned out to be a surprisingly strong predictor of conversion in your product?