The Number Goldman Stopped Using

If you have read anything about the creator economy in the last three years, you have read this number. It is in the opening paragraph of a thousand blog posts, on the first slide of every pitch deck, and in the Wikipedia entry — which cites it and nothing else.

Before going further, one correction to a claim we nearly published ourselves: "half a trillion" is not an exaggeration invented downstream. It is Goldman's own headline. Sites using that phrasing are quoting the bank accurately. The distortions in this chain are elsewhere, and they are worse.

The April 2023 article estimates the creator economy's total addressable market at $250 billion in 2023, roughly doubling to $480 billion by 2027. It counts three monetization channels: brand deals, a share of platform advertising revenue, and direct payments from audiences. It puts the global creator population at 50 million, growing at a 10–20% compound annual rate.

It does not publish a methodology. It does not break the $250 billion into components. A trade report on the underlying research note, thirteen days earlier, said the same thing in plainer terms — the note "does not specify what components make up the $250B figure."

So from the first day of its life, the most-cited number in this industry had no published composition. Everything that follows is people passing along a figure nobody could take apart.

The pattern is not what we expected. Distortion does not track commercial interest. The most rigorous source found in this walk was a marketing agency — which labelled the figure "one firm's projection, not consensus," linked the primary source, and noted that growth rates are more stable across firms than absolute dollar projections. The worst offenders were SEO listicles assembled by aggregating other listicles. The mechanism here is production method, not malice.

Two years after the original, Goldman published follow-up creator-economy research. It contains no dollar market-size projection at all.

The follow-up revises the creator population to roughly 67 million in 2025 — described as up from about 50 million in 2022, which quietly re-dates the original baseline by a year — and projects growth at about 10% annually to 107 million by 2030. That 10% is the bottom edge of the original's 10–20% range. The upper half of the assumption is gone.

The only market figure it defines precisely is narrow and an order of magnitude smaller: US influencer marketing at roughly $26 billion in 2023, about a third of US social digital ad spend. That is the one number in Goldman's creator-economy research with a stated denominator, and it is almost never quoted.

What is not in question is the shape: the bank published a number, never published its composition, then two years later published research that declines to produce a figure of that kind at all. The April 2023 page carries no correction, no update note, and no "as of" stamp. It still reads as current, and the industry still reads it that way.

Take the 2023 report entirely at face value — both halves of it. It gives a market size and, in the same breath, a creator count with a growth rate. Divide one by the other.

This is the whole trick, and it is not even concealed — both numbers are in the same article. A market doubling while its participant count doubles is not an opportunity getting bigger. It is an opportunity staying the same size and being shared out more thinly.

The second row settles the central question. If Goldman's figure described money creators actually receive, US creator income would need to be roughly twenty times what the most careful independent measurement finds. No plausible adjustment for the rest of the world closes a gap that size. Whatever the half-trillion is, it is overwhelmingly not money arriving in creators' accounts.

The fourth row is its own small scandal: two firms, the same year, the same nominal subject, and a figure roughly double the other. At least one of them is counting something the other isn't, and neither says which.

The unresolved ambiguity at the centre of all this is one phrase — "a share of advertising revenues with the host platform." Read one way it means the creator's cut. Read the other it means the gross platform advertising revenue that creator content attracts. On YouTube those differ by the whole 55/45 split. Goldman never says which it means, and that single unstated choice could account for most of the distance between $250 billion and $20 billion.

Five thousand dollars per creator is the average implied by Goldman's own two numbers. Against the distribution, that average is close to meaningless.

Goldman's research puts the share of creators earning $100,000 or more at about 4%. Separately — and this attribution matters, because the two figures are routinely welded together and credited to the bank — a 2022 survey by Linktree, not Goldman, found more than 70% of creators earned under $500 a year.

We nearly published those two figures as a single Goldman finding. They are not. One is a bank's estimate of the top of the distribution; the other is a link-in-bio company's survey of the bottom. They point the same way, which is exactly why it is tempting to merge them, and exactly why merging them would be the error this issue is about.

Taken separately, both hold. A market of $250 billion in which a majority of participants earn less than $500 a year is only coherent if the money is overwhelmingly not going to the participants. Which is, at last, an answer to what the half-trillion measures: it is the size of an industry built around creators, and it has never been a description of what creators get.