Two out of three influencers have fake followers. Here's how to spot them.
India is the largest source of fake-follower traffic in the world. Most agency audits miss the things that actually matter. Here's what we check before we sign a creator.
Over 68% of Indian Instagram creators artificially inflate their followers or engagement, according to HypeAuditor's audit of the market. Two out of three. India sits among the highest fake-follower rates in the world.
That's not a controversial take. That's the floor.
And it's expensive. When CHEQ and the University of Baltimore first put a number on it, influencer fraud was already costing brands roughly $1.3 billion a year, with about half of all engagement on sponsored content fake. It hasn't gotten cleaner since.
What brands keep getting wrong is what to do about it.
Three things that don't work
1. Filter by follower count. A 1M creator with 60% fake followers is worse than a 100K creator with 5% fake. The first costs more and delivers less. HypeAuditor's tier data makes the point: the 5K to 20K band, the sweet spot most brands shop in, shows growth anomalies in 59% of accounts, the highest of any tier.
2. Trust the engagement rate alone. Engagement is inflated by the same bot networks that inflate followers. With roughly half of sponsored-content engagement fake (CHEQ and University of Baltimore), a strong ER on its own tells you almost nothing. It's necessary, not sufficient.
3. Run a quick agency audit. Most "audits" check follower count, ER, and the last few posts. They miss audience quality entirely.
What actually works
- Audience demographics deep-dive. Where do followers actually live? What's the gender split? Most beauty creators with female-coded content have 70 to 80% male followers, quiet but devastating to campaign performance. We've watched mid-funnel beauty briefs collapse on this metric alone.
- Comment-quality scoring. Real audiences argue, ask questions, share stories. Bot audiences emoji-spam.
- Cross-platform overlap. A creator's audience on Instagram, YouTube, and Twitter should overlap meaningfully but not perfectly. Identical audience profiles across three platforms is a red flag.
- Historical campaign data. Did they actually drive sales, installs, or sign-ups in past campaigns, or just views?
The honest version
This is grunt-work. It's not a tool. It's a 90-minute session per creator before you sign them. The agencies that do this work consistently are the ones whose campaigns actually perform.
We've audited 3,200+ creators across 535+ campaigns. The pattern is the same every time: audience quality beats follower count, every single time.
