Follower reach is mostly phantom, so brands pay for the audience, then pay again to reach it. The $32 versus $21 CPM maths, worked honestly.
Influencer marketing has a pricing convention so familiar that almost nobody examines it. Fees are set on follower count. A creator with a million followers commands a multiple of one with a hundred thousand, and the logic feels like media logic, more followers, more reach, more value. The convention contains an assumption, that followers are reach, and the assumption stopped being true years ago. Working through what actually happens to a sponsored post, with the maths visible at every step, explains a budget anomaly many brand teams have noticed without being able to name, the influencer line keeps growing and the reach numbers keep not.
To be clear at the outset, this is an argument about using influencers as a reach channel, as advertising. Influencers used for what they are uniquely good at, credibility, buzz, social proof, the PR job, are a different investment judged on different terms, and the maths below does not damn them. It damns a substitution.
The fee buys the audience, priced per thousand followers. The first discount arrives immediately, because a follower count is a historical artefact, everyone who ever followed, including the dormant, the bots, the people who followed for a giveaway in 2021. The active audience is a subset.
The second discount is the one that breaks the convention. Organic delivery is algorithmic, platforms long ago stopped showing posts to everyone who followed an account, the feed is assembled by predicted interest, and a typical post reaches a single-digit to low-teens percentage of followers. Sponsored content tends to travel worse still, audiences engage less with visibly branded posts, and engagement is what feeds distribution. The million followers the fee was priced on resolves into something like fifty to a hundred and fifty thousand actual exposures. The brand has paid for an audience, and most of the audience never saw the post. That reach was phantom, priced as if it were real.
Now the campaign is live, the organic numbers arrive, and they are underwhelming, so the standard fix is applied, paid amplification. The post is boosted, put on media budget and delivered through the ad auction to the people it did not reach organically, which is to say, mostly the influencer's own followers, the audience the brand already paid for once. This is the double payment, fee for the audience, media to actually reach the audience, two invoices for one group of people.
Run the effective numbers and the anomaly becomes a figure. Take the full cost of reaching people this way, the fee amortised over real delivered exposure plus the boosting media, and it works out around a $32 effective CPM. The comparable cost of reaching the same people directly through the platform's own auction, with advertising the brand controls, sits around $21. The premium channel is the one with the phantom reach. The brand is paying roughly half as much again per thousand, for delivery it had to buy twice, carrying content it does not control, that encodes the creator's brand as much as its own.
As a reach mechanism, the structure is simply inefficient, and no negotiation fixes it, because the inefficiency is the structure. The fee prices an audience the platform will not deliver, the boost buys the delivery separately, and the brand sits at the end of the chain paying both invoices.
The resolution is not to cut influencers, it is to reprice what they are for. The fee was never really buying reach, even when everyone involved believed it was. What an influencer uniquely sells is belief, the trust of a specific audience in an independent voice, and belief is a PR asset, not a media one. Launching into a community, borrowing credibility a brand cannot self-generate, creating the social proof a buyer checks at the consideration moment, generating the buzz that earns attention beyond the audience itself, these are real effects, sometimes spectacularly valuable, and none of them are priced per thousand.
Judged as PR, the investment maths changes shape entirely. The question stops being what the CPM was and becomes what the belief was worth, the same way nobody evaluates a great piece of earned media on cost per impression. Some creator partnerships are bargains at any follower count on that test, others are expensive at any price, because no amount of audience fixes content nobody believes.
And judged as reach, the conclusion is equally clean, reach is what advertising is for. Systematic delivery into the whole buying population, light buyers included, at auction prices, with creative the brand controls, measured in attention and brand effects. Media creates the opportunity to see your ads. Creative determines whether they work. Both halves of that sentence belong to the brand, neither of them to an intermediary's follower count. The platforms sell the reach directly, the $21 was always available, what it requires is advertising good enough to earn its place in the feed, which is a creative problem, and a solvable one.
The practical exercise takes an afternoon. Pull the influencer line for the last year, split every engagement into the job it was really doing, belief or reach. Price the belief work as PR and keep what earns its place, there will be real keepers. Then take everything that was actually a reach buy, calculate its true effective CPM, fee plus boost over genuinely delivered exposure, and set it against the direct alternative. The gap is the double payment, and for most brands it is large enough to fund the native, brand-controlled creative that closes it.
Influencers deserve better than being an inefficient media channel, and brands deserve better than paying twice for an audience that was never really for sale. Pay for influence where influence is the job. Buy reach where reach is sold honestly, one invoice at a time.