We’ve been experimenting with something that’s been on my mind for a while: using AI to do the heavy lifting on creator-to-brand matching across markets. The idea is simple—instead of manually scrolling through thousands of profiles, AI can analyze creator content, audience, values, engagement patterns, and brand DNA, then surface the creators who actually fit.
But here’s what I’ve been wondering: is it actually smarter, or is it just making recommendations faster and calling it intelligence?
So far, we’ve run a small pilot where we used AI matching for a US beauty brand looking to expand into Russian social commerce. The AI pulled together about 50 creators across TikTok and Instagram—mix of macro and micro influencers—based on factors like audience overlap, content style alignment, engagement authenticity, and brand affinity signals.
Then we did the boring work: manually reviewed these 50, narrowed it down to 15, had actual conversations with them. The 15 we ultimately partnered with were genuinely strong fits. They understood the brand positioning, their audiences actually converted, and the content felt authentic.
But here’s the interesting part—some of the creators the AI flagged as “high match” turned out to be meh when we talked to them. They had the right metrics, but something was off in the conversation. And a few we partnered with weren’t even in the AI’s top 20, but the human conversation revealed they “got” the brand in a way the metrics didn’t capture.
I think the real value is the reduction in noise. Instead of evaluating 2,000 creators, we evaluate 50. That’s meaningful time savings. But the actual matchmaking—knowing whether a partnership will actually work—still requires human judgment.
Has anyone else tried this? I’m trying to figure out if we’re using AI as a filter (which seems to work well) or if it’s actually making smarter strategic decisions about fit. What’s your experience been?