Who Commissioned the Creator Economy Survey, and What That Changes

The Two Surveys Everyone Quotes

When journalists and policy briefs cite creator-economy statistics — the number of people who self-identify as creators, median reported incomes, the share treating content as a primary livelihood — two sources appear more than any others: a survey series commissioned by Adobe and a report published by Linktree. Both circulate as industry benchmarks. Neither is a government statistical product.

Adobe's creator-economy survey has appeared in successive waves since at least 2021. The company's stated methodology describes online panels recruited through third-party sample providers, with respondents self-identifying as "creators." The 2022 wave, widely cited in press coverage, reported a global figure of roughly 165 million people entering the creator economy since 2020 — a number Adobe derived by weighting self-reported survey responses, not by counting verified platform accounts or tax filings. Adobe sells Creative Cloud, a subscription software suite directly marketed to the creator segment the survey flatters. That commissioning relationship does not invalidate the methodology, but it is a condition of the research that independent readers cannot subtract from the result.

Linktree, a platform that provides link-in-bio tools to direct followers toward monetised content, published its Creator Report in 2022 in partnership with a market-research firm. The report cited more than 9,000 respondents across multiple countries. Linktree's own registered user base was a recruitment channel for the sample — a structural feature the report disclosed in its methodology section, though coverage rarely repeated it. When a platform surveys its own users about creator income, it is sampling people who have already decided to build an audience-linked presence, which skews the population toward those with higher engagement intent than the broader category of "anyone who posts content online" implies.

What the Methodology Disclosures Actually Say

Neither survey used probability sampling. Both relied on opt-in online panels or platform-recruited participants, which the research industry classifies as non-probability samples. Non-probability samples cannot support confidence intervals in the statistical sense and cannot be straightforwardly extrapolated to a defined population — in this case, "all creators" — without assumptions the surveys themselves acknowledge only in footnotes.

The Adobe survey weights its data by region and demographic characteristics, a standard adjustment. What it cannot adjust for is the self-selection of people willing to complete a survey about their creative work, who are plausibly more engaged with their creative identity than those who decline. The Linktree report's recruitment through its own user base introduced a related problem: Linktree users are, by definition, people who have already installed an additional layer of audience-monetisation infrastructure, which positions them somewhat further along the creator funnel than a randomly selected social-media user.

Reported income figures in both surveys are self-reported, uncorroborated by platform payment data, and not adjusted for the costs creators incur — equipment, software subscriptions, editing time — that would convert gross into net. The Federal Trade Commission's guidance on earnings claims requires that income representations reflect what participants typically earn, net of costs, when those representations are used to recruit. Survey figures reprinted in media coverage carry no such obligation, but the same logic applies to their interpretation.

Reading the Provenance Alongside the Finding

None of this makes the Adobe or Linktree figures fabricated. Large self-reported samples, consistently gathered, can reveal directional trends — growth in the number of people experimenting with content monetisation, a shift toward multiple platform presences, rising interest in direct-to-audience revenue — even when the absolute numbers are soft. The appropriate question is not whether to discard the data but whether to treat it as a census-quality count.

The commissioning relationship is a structural feature of how creator-economy research gets funded. Government statistical agencies — the Bureau of Labor Statistics, in its Current Population Survey multiple-jobholder data — produce narrower figures on secondary employment, but they do not ask respondents whether they consider themselves creators. The gap between what official statistics cover and what the industry wants to know is real, and it is the gap that commissioned surveys fill. That is worth noting plainly: the surveys exist because a data vacuum existed, and the parties best placed to fund research into that vacuum were the companies with a commercial interest in the answer.