When we were scaling Cal AI, I evaluated over 10,000 influencers (but we only signed 300 of them).

The most expensive mistake I watched other marketers make was always the same…

They'd spend hours studying a single creator, pouring over their social media analytics and debating it with the team for a week.

Here's what doing this at scale taught me: if it takes you that long to decide, the answer is already no.

You should be able to look at an influencer and know within 20 seconds whether they'll convert for your product.

There's a famous Supreme Court case where justice, Potter Stewart, was asked to define a what constitutes obscene content and he couldn't do it. He just said, "I know it when I see it."

That's exactly how I feel about magnetic influencers versus mediocre ones.

When it comes down to it, this is the 3-step process I use to instantly know whether or not we should sign someone:

#1: How many views do they get on average?

This is more about their baseline content metric, because anyone can go viral with enough content volume. I care about the floor, not the ceiling.

#2: What does the comment section look like?

This is by far the most important (but often the one that operators skip).

There's a huge difference between a comment section that's all 🔥🔥 💪🏻💪🏻 

and one that's full of "thank you for posting this, it changed my life."

The first audience is there to be inspired and entertained, while the second one will buy whatever that creator tells them to.

#3: Do they have a personality, and could you actually picture yourself being friends with this person?

Because that's what you're really paying for.

You're hiring the relationship that influencer has with their audience, not just their name and face.

So if the answer to any of those three questions is no, just walk away. I don't care how big they are.

This simple test comes down to pattern recognition, and the only way to learn it is to look at thousands of creators and be wrong on a bunch of them.

So if you're sitting there building a case for why a creator might work, stop.

The good ones never require a case, and with time (and this 3-step criteria) you’ll just know.

- Jake