For fifteen years, one number decided who got the brief. It sat next to every handle in every casting deck, and it settled arguments before they started: bigger reach, better buy. The follower count was influencer marketing's first metric — and, for most of its history, its only one.
It survived that long because it was easy, not because it was good. A follower count is a measure of accumulation. It tells you what a creator managed to gather, over years, across trends, through however many pivots in style and subject. It says almost nothing about what that creator makes now, who actually watches it, or whether any of it belongs anywhere near your brand.
Every experienced marketer knows this. It's why the best casting has always been done by feel — a brand manager scrolling a profile and simply knowing, within seconds, whether the creator fits. The problem was never that taste didn't work. The problem was that taste doesn't scale past the markets and scenes one person knows by heart.
A number that measures the wrong thing
Consider what the follower count actually rewards. Longevity. Early arrival on a platform. A viral moment two years ago. An audience gathered around content the creator may no longer make. None of these are qualities a brand buys when it books a collaboration. What it buys is the next ninety days of that creator's output, placed in front of the people who choose to watch it.
That gap — between what the number measures and what the brand buys — is where most wasted creator budget lives. The big-but-wrong creator is expensive twice: once in fee, and once in what it quietly does to the brand to appear in the wrong feed, in the wrong company, in the wrong tone.
The follower count is a measure of accumulation. The brand is buying the next ninety days.
What the content already knows
The alternative was always sitting in plain sight: the content itself. A creator's recent posts carry everything a casting decision needs — the aesthetic, the pacing, the products they reach for unprompted, the way their audience talks back to them. A human with taste reads all of this in a scroll. The shift happening now is that machines can finally read it too — image, video, audio, and text together, at the scale of whole markets rather than one profile at a time.
This is the idea behind our brand-fit work: learn what on-brand looks like from the creators a brand has already approved, then find the content-twins of those creators anywhere. Not "who is big in Germany" but "who in Germany makes the thing we already know works." The follower count doesn't disappear — it becomes what it always should have been, one input among many, somewhere behind fit.
What changes when you cast this way
Three things, in our experience running programmes this way:
- Smaller creators get booked more. When fit leads, the mid-sized creator whose content is precisely on-brand beats the large one whose content is nearly right. Budgets stretch further and the content that comes back needs less correcting.
- New markets stop being a gamble. Casting by feel ends at the border of the markets you know. Casting by content travels — the same standard, applied everywhere the brand goes.
- The brand stops drifting. Every off-brand booking erodes what the brand means a little. When the universe of castable creators is defined by fit, reach stops coming at the price of equity.
None of this makes taste obsolete. It makes taste the training data. The marketer's eye still decides what on-brand means — the system's job is to hold that standard steady across ten markets and two hundred thousand profiles, long after the deck with the follower counts has been retired.
