What marketers themselves say they are doing in 2026

The market-size models above are built by research firms. The demand-side survey is a separate instrument, and IMH’s 2026 edition is the most recent one published.

IMH’s Benchmark Report 2026 draws on 600+ respondents and reports that “72.22% expect their influencer budget to increase by 50%+”, with 87.49% expecting increases overall (Influencer Marketing Hub, Benchmark Report 2026).

From the same edition: 66.3% “report running programs entirely in-house”; 31% “of respondents included TikTok in their influencer plans”, making it the most-planned platform in the survey; 51.43% indicate expansion intent for nano creators; 36.67% use AI primarily for creator discovery and sourcing; 70.37% of budget-increasing brands say they are using social commerce; and 45.9% of those answering use promo or discount codes as their leading measurement approach (IMH Benchmark 2026).

Editorial opinion. The three most telling lines in that set are the ones nobody quotes. Two-thirds of programmes running entirely in-house means the agency layer is being disintermediated at the same time budgets grow. A plurality measuring with promo codes means the industry’s attribution problem is being solved with the bluntest instrument available. And AI’s first job inside creator marketing turns out to be finding creators, not replacing them.

IMH’s 2026 respondents also report an aggressive payback expectation: 65.9% expect payback within one month, “including 48.4% within 2 weeks” (IMH Benchmark 2026).

against your own account. A one-month payback expectation on a channel whose brand effects accumulate over quarters is a mismatch waiting to produce disappointment. Our read is that the expectation is inherited from paid social, where it is measurable, and applied to creator work, where it usually is not. Testable: compare your own creator cohort’s 30-day and 180-day contribution before you set the KPI, not after.

What US advertisers are doing with the money — the IAB block

This is the best-instrumented part of the market: a trade body surveying named advertisers, publishing its methodology and its full report.

Nearly half of ad buyers — 48% — now treat creators as a “must buy”, placing the channel third after social media and paid search (IAB, November 2025).

“Nearly three-in-four creator ad buyers are already using or planning to use AI within the next year” — and the use cases are assistive rather than substitutive: content editing (49%), creator briefs (46%), personalisation (45%) (IAB, November 2025). The IAB prints that finding as a proportion, not a percentage; we quote it as it stands rather than converting it into one.

At the same time, 95% of advertisers report concerns about using AI in creator marketing, with loss of authenticity and human connection the single largest worry (IAB, November 2025).

Creator marketing is still not a settled line item. The IAB finds brands funding it from multiple pools at once — 62% from social media budgets, 53% from creator-specific budgets, with general digital and general advertising close behind (multi-select, n=453) — and concludes that creator marketing “is not yet universally treated as its own established line item. Instead, it’s often carved out of broader digital, social, or general advertising pools, i.e., funding can be reactive, inconsistent, or fragmented across teams” (IAB report PDF).

Editorial opinion. Read those four facts as one sentence and the market’s shape is obvious: the channel is growing four times faster than media overall, AI is already inside the workflow of three in four buyers, almost all of those buyers are worried that AI erodes the exact quality that brought them to creators, and more than half of them still do not own a dedicated budget line to make the trade-off deliberately. That gap — between growth and governance — is the operating environment for anyone building AI-creator products, ourselves included.

Where the market is growing fastest

The single biggest defect in the earlier version of this article was a regional table that mixed three research firms’ incompatible global bases into one breakdown. Shares that do not sum, and dollar values implying three different world markets, are worse than no table. This section therefore uses one base — Mordor’s — and nothing else.

North America generated USD 10.74 billion of 2025 influencer marketing spend, 34.55% of the global total and the largest single region (Mordor Intelligence).

Asia-Pacific is the fastest-growing region in the same model, at a 33.90% CAGR for 2026–2031 (Mordor Intelligence).

Editorial note on what is missing. We removed the European country-level forecasts that appeared here previously — a 2026 figure for the UK, one for Germany, a French growth rate, and a European regional total. The research firm they were attributed to does not publish them on the cited page, and the forecast that page does carry is incompatible with them by an order of magnitude. No European breakdown appears in this article because we do not have one we can stand behind. We also do not publish a North American CAGR: neither firm we checked publishes one, so the frequently repeated claim that APAC grows “three to four times faster than North America” has no denominator and is not made here.

AI creator production is geographically agnostic in a way human creator rosters are not: the character does not hold a passport, live in one time zone, or speak one language natively. If that advantage is real, it should show up as a higher share of multi-market launches among AI-creator campaigns than among human-creator campaigns in the same category. We consider it a plausible structural advantage, not a measured one.

Which platforms hold the campaigns

Collabstr’s 2026 report, based on an analysis of 21,000 brand collaborations, finds that “platform-agnostic user-generated content (UGC) accounts for 35% of influencer marketing campaigns worldwide, surpassing TikTok (21%) and trailing only Instagram (40%)” (eMarketer, reporting Collabstr’s data; Collabstr 2026 Influencer Marketing Report).

In the same analysis, “TikTok-led campaigns fell 48% year over year”, while “UGC now drives 29% higher conversions than non-UGC formats” (eMarketer).

Editorial opinion. The apparent paradox — TikTok-led campaigns falling by half while TikTok remains the most-planned platform in IMH’s 2026 survey at 31% — resolves cleanly. Brands are not leaving TikTok; they are ending TikTok-only campaigns. The asset is commissioned once as platform-agnostic UGC and distributed through creators across Reels, TikTok and Shorts simultaneously. That reframes what a brand buys from a creator: not a placement on one platform, but a reusable asset with usage rights attached. It also explains why the biggest single format in the market now has no platform name at all.

What share of budget goes to creators

IMH’s March 2025 edition found that 75.6% of marketers planned a dedicated influencer marketing budget for 2025, down from 85.8% in 2024 — a 10.2-point fall (IMH Benchmark 2025, archived PDF).

From the same edition’s budget-allocation chart: 14.4% of marketers put 10–15% of total marketing budget into influencer marketing, 12.7% put 5–10% in and 5.1% put under 5% in (IMH Benchmark 2025, archived PDF). Those are the chart’s own buckets, quoted individually — IMH does not print a combined “under 10%” figure, so neither do we. These are IMH’s numbers, not Statista’s and not the IAB’s — an attribution we corrected in this rewrite.

Editorial opinion. A falling count of dedicated budget lines alongside rising total spend is not a contraction; it reads as consolidation. Creator money is moving inside general social and performance budgets — which is exactly the fragmentation the IAB names — and the practical consequence is that the channel’s spend is increasingly hard to see in a plan even as it grows.

On the agency side, WARC’s research with impact.com found use of specialist agencies to find influencers had “risen to nearly 74%, up 20% from 2019”, and that “79% of brands use a suite of partners” (WARC). Two qualifications: this is the 2022/23 research wave, the most recent published, and WARC’s wording is “up 20%” — a relative rise, not 20 percentage points. Set against IMH’s 2026 finding that two-thirds of programmes now run entirely in-house, the direction of travel since that wave is worth treating as an open question rather than a trend line.

The channel is also being recognised in national ad-spend accounting. The AA/WARC Expenditure Report — which recorded £46.7 billion of UK media investment in 2025, up 6.4% — has begun restructuring its categories, and “a second phase will see a working group of stakeholders convened to address further areas for refinement, including developing a clearer understanding of investment in the influencer / creator channel” (Advertising Association, 30 April 2026). That is a UK market report, not a global one — but a channel getting its own line in national expenditure accounting is a maturity signal that no market forecast provides.

What ROI influencer marketing delivers — the honest answer

This is the most contaminated node in the entire influencer marketing literature, and the honest answer is uncomfortable: there is no independently audited ROI benchmark for this channel.

Editorial note on what is missing. The single most-quoted figure in the industry — a headline “return per dollar spent” that appears in hundreds of vendor pages, agency blogs and decks — does not have a retrievable primary source behind it in its current form. Tracing it back leads to a benchmark published years ago, measuring earned media value per dollar rather than revenue per dollar, restated ever since without re-measurement. We removed it from this article rather than repeat it with a hedge, and we removed the companion claims (“top campaigns return $X”, “N% of businesses double their money”) that travel with it from the same lineage.

The often-cited claim that influencer marketing generates “11× more ROI than traditional digital advertising” traces back to old Tapinfluence/Nielsen material and circulates in current sources without fresh verification. It is usable as a narrative argument about direction; it is not usable as a citable fact, and it does not belong in a business case.

So what do you plan against? Three things that are actually measurable:

1. Your own before-and-after. The IMH 2026 survey shows the industry’s own leading measurement method is promo and discount codes, used by 45.9% of those answering. It is blunt, but it is yours, and it is not a vendor’s average.
2. The channel’s growth rate rather than its return rate. The IAB’s +26% for 2025 and the 4× gap to overall media growth are measured, sourced and stable across editions. Growing four times faster than the medium it sits in is a real signal about where attention is going, even without a return multiple attached.
3. Payback discipline before the campaign, not after. If the industry expectation really is payback inside a month for two-thirds of buyers, the number to agree with your finance partner is the observation window — before the brief goes out.

The macro picture — where the creator economy is heading

Goldman Sachs Research estimated that the creator economy “could roughly double in size over the next five years to $480 billion by 2027 from $250 billion today”, that there are about 50 million creators globally, that “only about 4% of global creators are deemed professionals, meaning they pull in more than $100,000 a year”, that the creator population should grow at a 10–20% CAGR over five years, and that “brand deals are the main source of revenue at about 70%” (Goldman Sachs Research).

Hedge that travels with those numbers. This is an April 2023 estimate. It remains the most-cited macro sizing of the creator economy, and Goldman Sachs has not publicly restated it since — while its horizon, 2027, now arrives next year. Treat it as the best available macro frame and as a three-year-old one, in the same breath.

Editorial opinion. The most consequential line in that set is the 4%. When the industry says “50 million creators,” it is describing a population in which the professional tier is a rounding error and brand deals supply seven dollars in ten of what creators earn. For a brand, that means two things at once: a long, cheap shelf of nano and micro creators at the bottom of the funnel, and a very thin layer of people who can be relied on as professional partners at the top. It is also why creator-income tables should be read with suspicion — we removed ours in this rewrite, for reasons given below.

How much the industry loses to fraud

The one rigorous, independently reported sizing of influencer fraud remains CHEQ’s research with the University of Baltimore: influencer fraud “is expected to cost businesses $1.3 billion this year, according to new research from cybersecurity firm Cheq” (CBS News, 25 July 2019; original CHEQ / Roberto Cavazos study). No comparably rigorous update has been published since. That is a genuine gap in the industry’s knowledge, and it is the honest headline of this section.

Editorial note on what is missing. Current fraud figures circulating in the trade press — a 2026 industry-loss total, a split between bot-driven and AI-synthetic fraud, platform-by-platform and tier-by-tier fraud rates — trace back to two kinds of source: a named “2026 influencer fraud loss report” that we could not locate outside a closed circle of aggregator blogs citing one another (no publisher page, no methodology, no document), and self-published studies by vendors whose commercial interest is in the problem they are measuring. Neither clears the bar for a number printed on this site. All of it was removed. We would rather publish a seven-year-old figure with a named methodology than a current-sounding one with none.

What is measurable is what buyers report experiencing. In IMH’s 2026 survey, “fake or bot followers account for 56.5% of all reported fraud/quality issues” — the dominant category of the problems marketers say they actually hit (IMH Benchmark 2026). And the IAB’s advertisers name audience authentication standards among the capabilities they need to connect creator spend to outcomes (IAB report PDF).

Detection tooling built for bot-farm signatures — follower-to-following ratios, engagement spikes, activity timing — may generalise poorly to AI-generated personas that post consistently, hold conversations and behave like accounts run by people. This is an open question about a technical limit, not a tested finding, and we name no vendor as failing it: none has been independently benchmarked on the problem in public.

Forecasts to 2032 — and what they are worth

Table 2. Published forecasts, with their scopes

How to read that table. The Gartner prediction forecasts technology adoption, not market size — it belongs in a different column of your plan than a dollar forecast. The SNS Insider virtual-influencer forecast is a single-vendor model with a seven-year horizon and a 2023 base year; a 41% compound rate sustained for nine years is a strong assumption, and it sizes a market that is not a subset of the influencer marketing spend in Table 1 on the same definitions. We include it because it is the only primary-sourced sizing of that segment we could retrieve, and we flag its fragility for the same reason.

Three vectors we think are real

1. Platform convergence. Meta Advantage+ and TikTok Smart+/Symphony have converged on the same operating model: the advertiser supplies goal, budget and assets, and the platform does audience, placement, bid and creative variation. Google’s AI Max belongs in the same direction of travel but is a narrower object — a feature suite inside existing Search campaigns rather than a separate buying model. Google’s own description: “AI Max brings the best of Google AI into your Search campaigns with a suite of targeting and creative enhancements”, letting advertisers “expand your keywords to find untapped, performant queries with broad match and keywordless technology” (Google). The competitive edge moves from campaign hand-tuning to the quality of the strategic input, the depth of first-party data and the diversity of creative assets — which is precisely where creator content becomes a supply problem rather than a media problem. We cover the mechanics in our piece on AI-native advertising.

2. Authentication becomes its own line item. The IAB’s 95% concern figure and its advertisers’ named need for “standards for audience authentication and fraud prevention” point the same way from the demand side (IAB report PDF). Editorial opinion. A provenance and verification sub-market — C2PA credentials, watermarking, creator identity verification, deepfake detection — is being funded out of influencer budgets that previously bought only media and content. We deliberately attach no share-of-budget forecast to that: the prediction most often quoted for it is one we could not retrieve from a publisher-hosted page (see the removals table).

3. Regional weighting shifts toward Asia-Pacific. Mordor’s 33.90% APAC CAGR is the highest regional rate in its model. Editorial opinion. This is where geographically agnostic production has the clearest structural advantage — one character, localised into n markets, against human rosters bound to one city and one language.

Editorial note on the vector we removed. A fourth vector appeared in the earlier version of this article and has been deleted: a claim that CMOs forecast roughly 30% of influencer budgets going to virtual creators by 2026, converted into a dollar figure against the global market. The page carrying that figure labels it an “aspirational milestone target” — not a forecast — and quotes a materially lower number from the survey it cites. We could not retrieve that survey from its publisher, only from aggregators restating it, so neither figure appears here in any form. The derived dollar conversion was worse than unsourced: it implied more money flowing to virtual creators than the entire virtual-creator market was sized at by the same article’s other sources. When a number fits an argument that well, it deserves to be checked first rather than last. That is the rule we now apply to our own drafts.

What we removed from this article, and why

Publishing a list like this is unusual. We think it is the only honest way to update a piece that previously promised a source next to every figure and did not deliver one.

Methodological caveat

Direct comparison across the sources in this article is incorrect without context, and the differences are structural rather than accidental:

• IMH publishes marketer-survey data. Its 2026 edition draws on 600+ respondents; its 2025 edition does not disclose a sample size.
• The IAB surveys US advertisers and publishes the instrument alongside the results; the budget-source question cited above is multi-select at n=453.
• Mordor Intelligence builds a market model of the full ecosystem, with its own global base.
• Statista aggregates and restates other series; treat it as a mirror, not a measurement.
• Goldman Sachs sized the whole creator economy in April 2023, including monetisation streams that have nothing to do with brand campaigns.
• Gartner publishes predictions about behaviour and technology adoption, grounded in its own client surveys — not market sizings. We cite only predictions we could open on a Gartner-hosted page.

When you see “$X billion” in someone’s deck, the only safe way to use it is to name the source, the year and the scope. This article is built on that logic, and the previous section lists what happened to the figures that could not survive it.

What this means for a brand today

Influencer marketing in 2026 is an advertising channel on the scale of paid search or programmatic. In the one market where a trade body measures it directly — the US, per the IAB — it is growing about four times faster than the media industry around it. Three transitions are running through it in parallel:

1. Consolidation under AI-native platform stacks. The buying layer is automating end to end, and the scarce input becomes creative supply — which is a production question, not a media-planning one. Related reading: what an AI creator integration actually costs.
2. Format-first, platform-second commissioning. Platform-agnostic UGC is now the second-largest campaign type worldwide and TikTok-only campaigns have halved year on year. What you buy from a creator is shifting from a placement to a reusable asset with rights attached.
3. Authenticity becoming a budget line. Advertisers say at 95% that they are worried about AI in creator marketing, and they name audience authentication standards among the capabilities they still need. Disclosure, provenance and identity verification are moving from compliance overhead to line items — a direction we can source, at a pace we cannot.

Editorial opinion. A brand entering 2026 with “let’s book a couple of integrations and see” is operating off the 2018 map of this industry. A brand that sees the three transitions together can act on them cheaply, because none of them requires a big commitment to test: one platform-agnostic asset set, one properly disclosed AI-creator pilot in the market where localisation costs the most, one measurement window agreed before the brief goes out. If you are choosing where to start, our framework for picking a creator covers the selection half of that decision.

→ Browse the AI character catalogue

Frequently asked questions

What is the actual size of the influencer marketing market in 2026?

It depends entirely on scope, and three credible answers coexist. Mordor Intelligence sizes the global full ecosystem — brand–creator contracts, agency commissions, platform services — at $40.51 billion for 2026, up from $31.07 billion in 2025, and forecasts $152.56 billion by 2031 at a 30.36% CAGR. The IAB puts US creator ad spend alone at $43.9 billion for 2026, on a broader definition that adds paid amplification and branded content production, after $37.1 billion in 2025. Influencer Marketing Hub’s March 2025 edition projected $32.55 billion globally for 2025 from survey data; its 2026 edition contains no market-size figure at all. Comparing these figures without naming the scope is the most common error in influencer marketing decks.

Why do market estimates from different sources diverge?

Different definitions, not different realities. A US-only figure counts one country but includes paid amplification and branded content production. A global full-ecosystem model counts brand contracts, agency commissions and platform services worldwide. A marketer-survey projection estimates brand spend from what marketers report planning. And a separate, much smaller market exists for influencer marketing platform software — the SaaS tools brands use to run campaigns — which is frequently quoted as if it were the whole channel. Comparison only makes sense between figures on the same definition, in the same year, from the same firm.

What is the average ROI of influencer marketing in 2026?

There is no independently audited ROI benchmark for this channel, and we do not publish one. The most-quoted return-per-dollar figure in the industry has no retrievable primary source in its current form: it traces back to an earlier-decade benchmark measuring earned media value per dollar rather than revenue, restated ever since without re-measurement. The frequently cited claim that influencer marketing returns “11× more than traditional digital advertising” traces to old Tapinfluence/Nielsen material and circulates without fresh verification — usable as a narrative argument, not as a citable fact. What is measurable: your own before-and-after, the channel’s growth rate (the IAB’s +26% for 2025, about four times faster than overall media at +5.7%), and a payback window agreed before the campaign rather than after.

How much does the industry lose to fraud?

The last rigorous public sizing is CHEQ’s research with the University of Baltimore, reported by CBS News in July 2019 at $1.3 billion a year, and no comparably rigorous update has been published since. Current-sounding fraud totals circulating in the trade press trace either to a report we could not locate outside a closed circle of aggregator blogs, or to self-published vendor studies with a commercial interest in the finding, so we do not carry them. What buyers report experiencing is measurable: in Influencer Marketing Hub’s 2026 survey, fake or bot followers account for 56.5% of all reported fraud and quality issues.

Which region is growing fastest?

Asia-Pacific, at a 33.90% CAGR for 2026–2031 in Mordor Intelligence’s model — the highest regional rate it publishes. North America remains the largest region in absolute terms at $10.74 billion in 2025, 34.55% of the global total. We do not publish a North American growth rate or a European country breakdown: neither is available from the firms concerned on the pages we could verify, so the common claim that APAC grows “three to four times faster than North America” has no published denominator behind it.

What share of marketing budget goes to creators at a typical brand?

Fragmented rather than typical. The IAB finds brands funding creator marketing from several pools at once — 62% from social media budgets, 53% from creator-specific ones (multi-select, n=453) — and concludes that creator marketing “is not yet universally treated as its own established line item.” On the survey side, Influencer Marketing Hub’s March 2025 edition found 75.6% of marketers planning a dedicated influencer budget for 2025, down 10.2 points from 85.8% in 2024, with 14.4% allocating 10–15% of total marketing budget to the channel, 12.7% allocating 5–10% and 5.1% allocating under 5%.

Where is the market heading by 2030 and beyond?

The published forecasts agree on direction and differ on scope. Mordor Intelligence projects $152.56 billion globally by 2031 at a 30.36% CAGR. Goldman Sachs estimated the wider creator economy at around $480 billion by 2027 — an April 2023 estimate the firm has not publicly restated since. Gartner predicts that 60% of brands will use agentic AI for one-to-one interactions by 2028. SNS Insider forecasts a separate virtual influencer market at $154.6 billion by 2032 at a 41.29% CAGR from a $6.9 billion base in 2023 — a single-vendor model over a nine-year horizon, and a different market from the influencer spend above.

Why does this article list what it removed?

Because an earlier version promised a source next to every figure and did not deliver one. An adversarial fact-check found that several widely repeated statistics in it were attributed to pages that do not contain them, and that one report cited as a source could not be located outside a circle of aggregator blogs citing each other. Removing those figures quietly would leave readers unable to tell which parts changed. The removals table names each one and the reason, so anyone who quoted the earlier version can correct their own deck.