Why Follower Count Is a Terrible Way to Vet Creators

If you're running influencer campaigns, there's a good chance you've picked creators based on follower count at some point. It's the most visible metric. It's easy to compare. And it feels like a reasonable proxy for reach.

But follower count is a terrible way to vet creators. Here's why.

Followers can be bought

The influencer fraud problem is bigger than most brands realise. Studies suggest that anywhere from 10-25% of influencer followers are fake - bots, inactive accounts, or purchased followers designed to inflate perceived influence.

A creator with 500k followers might have real reach of 50k. You'd never know from the headline number.

Engagement rate is also gameable

"But I look at engagement rate too!" you might say. That's better - but still not enough.

Engagement pods, comment farms, and bot-driven likes are all common tactics. A creator can have a 5% engagement rate that's almost entirely artificial.

What matters is engagement quality:

• Are the comments meaningful or generic?
• Do the same accounts comment on every post?
• Is there genuine conversation, or just emoji spam?

What actually predicts campaign success

Based on what we've seen work, here are the signals that matter more than follower count:

1. Audience authenticity - What percentage of followers are real, active accounts?
2. Engagement quality - Are comments genuine and varied?
3. Content consistency - Does the creator post reliably in your niche?
4. Brand safety - Is there anything in their content history that could backfire?
5. Sponsorship history - Have they worked with competitors? Do they disclose properly?
6. Rate reasonableness - Is what they're charging in line with their actual value?

The bottom line

Follower count tells you almost nothing about whether a creator will drive results for your brand. It's a vanity metric that's easy to game and hard to trust.

If you're spending real budget on creator partnerships, you need to look deeper.

PlutoBa helps brands vet creators using structured signals - not vanity metrics. Request early access to try it.