What a Billion Reel Views Actually Buy: The Research on Food Creators and Restaurant Demand

Original AI-generated illustration for Flonci. Conceptual artwork, not a product screenshot.

A Times lunch with Topjaw's Jesse Burgess frames the dream version of food influence: a billion people watch the reels, and the creator ends up with Ferraris and a TV series made by Gordon Ramsay's company. The more useful question for anyone spending money is what those reels do to the restaurants inside them, and the research on that question is more specific than the headline.

Some stories travel because they confirm what an industry already wants to believe. The influencer version goes like this: a charismatic creator points a camera at a plate, the internet does its thing, and a restaurant that was empty on Tuesday has a queue on Saturday. The Times ran a lunch interview with Jesse Burgess of Topjaw under exactly that framing, with a standfirst noting that about a billion people watch his reels and that, as a result, he buys Ferraris and makes television with Gordon Ramsay's production company. The Telegraph has described Burgess and his co-founder Will Warr as the men who can make any restaurant an overnight success.

Only the headline and introduction of the Times piece are visible without a subscription, and this article will not pretend otherwise. But the visible framing is enough, because the framing is the thing worth examining. A billion views is a real number. A Ferrari is a real car. The implied middle step, the one where the views become a restaurant's revenue, is where the interesting research lives. Over the past two years, several peer-reviewed studies have asked precisely how food creators move people from watching to visiting, and at least one classic study measured what online reputation does to restaurant revenue with actual tax data rather than self-report.

Read together, they tell a story that is more useful than the headline and less comfortable than the pitch deck.

The Topjaw file, from the public record

Start with what is actually known about the case, because the case is genuinely instructive. Topjaw was co-founded by Burgess and producer Will Warr and has been on YouTube since 2015, according to an interview Burgess gave EE72 this month. The channel spent years making longform city guides, nearly died when attention moved to TikTok, Shorts and Reels in 2020, and then found its format: short, energetic interviews and reviews built around chefs and restaurants. When the first Best of London interview went up, Topjaw had 70,000 Instagram followers. It got, in Burgess's words, well into the millions of views. The operation now counts over half a million YouTube subscribers and a million Instagram followers, has filmed in well over 3,000 restaurants, and has hosted guests from Kate Winslet to Angela Hartnett. Burgess also hosts Knife Edge: Chasing Michelin Stars on Apple TV, made by Ramsay's production company.

One detail in that interview deserves more attention than the follower counts. Burgess says he will never take money from a restaurant. His ambition, in his own words, is for Topjaw to be the most authentic and trusted resource in food discovery in the world. Keep that detail in mind. It turns out to be the whole mechanism.

Two ways to read "overnight success"

When a venue claims a creator made it an overnight success, two very different claims are available. The weak version is that the video made people want to visit. The strong version is that the video produced revenue the restaurant would not otherwise have earned. These are not the same claim, they are not measured the same way, and the distance between them is where marketing budgets go to be misunderstood.

Most of the serious research on food influence measures the weak version, intention, and measures it well. A smaller body of work measures the strong version, actual demand, and it mostly does not study influencers at all. Honest analysis has to keep the two separate: use the intention research to explain the mechanism, use the demand research to bound the expectation, and refuse to quote one as if it were the other.

The intention engine, part one: trust is the bridge

The most directly relevant study is "From Feed to Table: The Role of Food Influencers in Restaurant Choices," published in the journal Tourism and Hospitality in March 2026. The researchers surveyed 474 social media users in Ecuador, screened so that every respondent actually follows at least one food influencer, and tested a model built on the Theory of Planned Behavior with one addition: trust in the influencer, included both as a direct antecedent of visit intention and as the mediating link between a follower's identification with the creator and their intention to visit.

The method is worth respecting because the paper is transparent about it. The sample was a non-probability convenience sample recruited online. The measures were validated scales from prior research, the model was estimated with partial least squares structural equation modeling with 5,000 bootstrap resamples, and predictive performance was checked with PLSpredict rather than just fit statistics. The model explains about half the variance in visit intention, which for survey research of this kind is respectable.

The finding that matters: identification with the creator, the parasocial pull that makes a viewer feel they know Jesse Burgess, translates into stated visit intention only when it passes through perceived credibility. Attitude toward the recommendation and the viewer's sense of control over the decision did the heaviest lifting; raw identification alone did not. In plain terms, the audience's feeling of closeness is not the active ingredient. The audience's belief that the creator is telling them the truth is.

The intention engine, part two: the content has to earn the relationship

A second study, "Virtual Culinary Influence," published in January 2025, published in the journal Information, approached the same question from the content side. Using the stimulus-organism-response framework, the researchers surveyed 347 MTurk workers, of whom 320 valid responses were analyzed, and asked what about food review vlogs drives engagement, parasocial relationships, attitudes and visit intentions.

Three content qualities did the work on engagement: informativeness, entertainment and vividness. On the creator side, attractiveness and homophily, the viewer's sense that this person is like me, were the significant predictors of parasocial relationship strength. Engagement and parasocial bonds then fed attitude and visit intention, consistent with the first study's chain. The study's limitations are plainly stated: an MTurk sample skewing male, Caucasian and Asian, self-reported behavior, and respondents who were largely not fans of any particular vlogger.

Here is where it gets interesting. In this second study, creator credibility did not significantly predict parasocial relationships, which on its face contradicts the first study's trust-as-bridge finding. The authors note the credibility measure itself was weak, with items eliminated during validation, and that a sample of non-fans may simply not contain the trust dynamics that exist between a creator and an actual audience.

The tension is the finding

A careless summary would average the two studies into mush. The better reading respects the disagreement, because the disagreement is informative. In a sample of people who actually follow food influencers, screened for exactly that, trust was the bridge between identification and intention. In a sample of general internet workers watching vloggers they mostly do not follow, credibility did not predict the relationship, partly because there was no real relationship to measure.

The synthesis that survives both: food-creator influence on visit intention is not a broadcast effect that lands on strangers. It is a relationship effect that operates on an accumulated audience, and its active ingredient is the audience's belief that the recommendation is honest. Reach determines how many people are inside the relationship. Trust determines whether the relationship converts to intention at all.

This is, not coincidentally, exactly how Burgess describes his own operation. The content qualities the second study found, informativeness, entertainment, vividness, are a fair description of the Topjaw format. The trust asset the first study found is the thing Burgess says he protects by refusing restaurant money. The man has not read the Ecuadorian structural equation model. He does not need to. He is running it.

From intention to revenue: the evidence gets thinner and more interesting

Intentions are not revenue, and the research community knows it. The strongest measured evidence on online reputation and restaurant demand predates TikTok entirely. Michael Luca's "Reviews, Reputation, and Revenue: The Case of Yelp.com," published in the Economic Journal in 2016, combined Yelp ratings with Washington State Department of Revenue data and exploited Yelp's rounding thresholds for a regression discontinuity design, about as close to causal identification as observational marketing data gets. The headline result: a one-star increase in Yelp rating leads to a 5 to 9 percent increase in revenue. Two details matter for our purposes. The effect was driven by independent restaurants; chains showed no measurable effect. And it measures the platform's aggregated reputation, not any single reviewer's influence.

On the creator side specifically, the most quoted numbers are survey self-reports. An MGH survey of 1,139 U.S. TikTok users in 2022 found that 38% of respondents, and 53% of millennial respondents, said they had visited a restaurant after seeing it on the platform, with 72% citing the visual appearance of the food as the trigger. These are real numbers from a named survey with a stated margin of error, but they are an agency's survey of recalled behavior, not measured transactions, and "visited after seeing" is not "visited because of."

The honest summary of the demand evidence: the mechanism from creator content to visit intention is well documented and runs through trust. The causal effect of online reputation on restaurant revenue is real, measurable and roughly single-digit percentages per star for independents. The specific causal claim that any given viral video produced any given queue is almost never measured at all, because measuring it requires transaction data, a baseline, and a counterfactual, none of which fit in a press release.

Why the man will not take the restaurant's money

Back to the most telling detail in the Topjaw story. Burgess's refusal to take money from restaurants is usually told as a charming bit of integrity. It is more precise to read it as asset management. The research says the asset that converts audience into intention is perceived credibility. A creator who sells reviews is selling the very thing that makes the reviews worth buying, and the audience eventually prices that in.

Notice where the money in the story actually comes from. Not from the restaurants in the videos. From the audience's attention itself: platform revenue, production, and a television career built on the credibility the videos accumulated. The Ferrari is paid for by the trust, and the trust is protected by the refusal. This is not a morality tale. It is a monetization architecture, and it generalizes: the durable creator businesses are the ones whose revenue does not require compromising the mechanism that creates the revenue.

For a brand or a restaurant group, the implication cuts both ways. A creator whose influence runs on perceived independence is the most valuable partner and the least purchasable one. The offers that would be easiest to make are the ones that would destroy the value being bought.

What a venue or brand should actually do with this

The research supports a short, unglamorous operating list.

First, buy the relationship, not the reach. The intention studies agree that conversion runs through an accumulated, trusting audience, so a smaller creator whose followers actually act on recommendations will usually outperform a bigger one whose audience is rented attention.

Second, measure the strong claim, not the weak one. Views and even visit-intent lifts are weak-version evidence. The closest a venue can get to Luca's standard is a tracked baseline: reservations, covers, or sales in the weeks before and after a creator feature, compared with a similar period, with the creator's audience geography in mind. A spike that decays in two weeks is a different asset than a step change in baseline demand, and only measurement can tell them apart.

Third, respect the capacity constraint. A restaurant is not an app. The demand a creator can generate in a weekend exceeds the seats a venue has, so the measurable upside is bounded by capacity while the reputational downside of a bad visit is not. Overflow demand is real but it spills to whoever is next on the list.

Fourth, do not try to buy the review. If the trust mechanism is the engine, pay-for-praise is sand in it. Sponsored formats have their place, clearly disclosed, but a venue that wants the Topjaw effect cannot order it off a rate card, because the effect is precisely what the creator's independence produces.

Where Flonci fits

This blog exists to help marketing decision-makers replace influencer folklore with evidence, and the food vertical is the folklore's natural habitat. Flonci's approach to creator selection starts from exactly the distinction this article drew: audience quality and credibility signals over raw reach, and measurement built around outcomes a business can verify rather than attention a platform can count. For brands evaluating food and hospitality creators, that means looking past the view count to the relationship underneath it, and setting up the measurement before the collaboration, not after the queue forms or fails to. The research says trust is the bridge. The practical work is knowing, creator by creator, whether the bridge is there.

Conclusion

The Topjaw story is true, and it is also not the model. A billion views can make a creator rich and famous, and can make individual restaurants very busy, but the mechanism is not the views. The mechanism is a trusted relationship between a creator and an audience, built video by video, protected by refusals nobody sees, and converted into visits only when the audience believes the recommendation. The intention research documents that mechanism in detail. The revenue research says online reputation moves independent restaurant sales by real, measurable, single-digit percentages, not miracles. And the one measured thing nobody can sell you is the thing the best creators refuse to sell themselves.

The next time a headline promises that a creator can make any restaurant an overnight success, the informed reading is: sometimes, briefly, through trust, within capacity, and almost never in a way that survives being bought. That is not a smaller claim than the headline. It is a usable one.

Research notes and sources

This article was prompted by The Times' lunch interview with Jesse Burgess ("The Horses review," October 2026). Only the headline and standfirst of that article are publicly visible without a subscription; the standfirst's remark about a billion reel views, Ferraris and the Gordon Ramsay TV series is quoted from that visible introduction, and nothing beyond it is attributed to the piece. Biographical and business facts about Topjaw are drawn from Burgess's interview with EE72 (October 1, 2026), including the 2015 founding with Will Warr, the 70,000-follower inflection point, the 500,000-plus YouTube and one million Instagram counts, the 3,000-plus restaurants filmed, the Apple TV series made by Ramsay's production company, and Burgess's stated refusal to take money from restaurants.

The two intention studies are peer-reviewed and open access: "From Feed to Table: The Role of Food Influencers in Restaurant Choices," Tourism and Hospitality (2026), a survey of 474 Ecuadorian social media users screened for following at least one food influencer, analyzed with PLS-SEM; and "Virtual Culinary Influence: Investigating the Impact of Food Vlogs on Viewer Attitude and Visit Intention," Information (2025), an MTurk survey of 320 valid respondents analyzed with SEM. Both measure stated intention, not observed behavior, and both rely on non-probability, self-reported samples; the second's authors flag the MTurk sample and the weak credibility measure as limitations, and that study's non-significant credibility finding is reported here rather than smoothed over.

The measured-revenue evidence is Michael Luca's "Reviews, Reputation, and Revenue: The Case of Yelp.com," Economic Journal (2016), described here: a regression discontinuity study using Washington State revenue data, finding a 5 to 9 percent revenue increase per additional Yelp star, driven by independent restaurants. It measures aggregate platform reputation, not individual creators, and predates short-form video. The TikTok visitation figures come from an MGH survey of 1,139 U.S. TikTok users as reported by Restaurant Dive (2022): agency-conducted, self-reported, with "visited after seeing" not establishing causation. Nothing in this article is a verified FLONCI client result or a commitment to an outcome. The cover image is an original AI-generated illustration, not a photograph of any person or venue mentioned.

When Creator Marketing Becomes Infrastructure, the Scoreboard Changes

Original AI-generated illustration for Flonci. Conceptual artwork, not a product screenshot.

The fourth annual LTK-Northwestern creator marketing study, published on October 21, 2026, announces that creator marketing has entered its infrastructure era. The headline is about scale. The more useful news is in what brands now count as success.

Once a year, the creator economy gets a measurement of itself that is worth reading slowly. The LTK and Northwestern University Creator Marketing Study is now in its fourth edition, and this year's release comes with a confident title: creator marketing has moved beyond one-off campaigns and become always-on marketing infrastructure.

The supporting numbers are real numbers. Ninety-one percent of brands in the study operate creator programs with an always-on component. The average number of creators brands work with rose 275% year over year. Creator programs now reach across marketing (74%), social (65%), media (61%) and brand teams (60%), with PR and affiliate teams increasing their involvement as well.

A headline like that tends to travel as a victory lap. It is more useful as a diagnostic. When a channel becomes infrastructure, the questions that matter change: not whether to use it, but how it is measured, who inside the company owns the decision, and what the measurement can actually support. This article reads the study the way a marketing decision-maker should read any vendor-partnered research: with respect for what it documents, and with care about what its numbers can and cannot carry.

What this study actually is

Before the findings, the instrument. The 2026 study was conducted by LTK in collaboration with Northwestern University in June 2026. It surveyed 209 C-suite marketers and senior decision-makers at brands that invest in creator marketing, the majority at organizations with more than $100 million in revenue.

Three properties of that design deserve to be stated plainly, because they shape every conclusion that follows.

First, this is a survey of believers. Every respondent works at a brand that already invests in creator marketing. The study cannot tell you whether creator marketing works in general; it tells you what experienced practitioners inside large, committed organizations report doing and valuing. That is still valuable, but it is a different claim.

Second, LTK is not a neutral observer. It is a creator commerce platform that sells exactly the kind of infrastructure the study recommends, and says so in the same release: more than 4,000 brands use its platform, and it used the occasion to introduce a new agentic campaign experience. The university partnership adds methodological credibility, and the release is unusually transparent about its sample. Neither fact removes the commercial interest. A reader should hold both at once.

Third, the year-over-year comparisons are directional, by the study's own admission. The 2026 survey updated its sample and questionnaire, so the dramatic movements, including the 275% rise in average creator counts, should be read as a change in what the study measured as much as a change in the market. Directional does not mean invented. It means a reader should not build a budget on the third digit.

None of this makes the study dismissible. It makes it readable. The right posture is to take its internal details seriously, especially the ones that are inconvenient for a simple growth story, and to discount its headline metaphors accordingly.

The scoreboard moved, and that is the real story

The most consequential finding in the 2026 study is not the always-on headline. It is the shift in what brands say they measure.

Asked to name the most important reporting metrics for creator campaigns, marketers put awareness and campaign sales in a tie at 51%, with conversion rates immediately behind at 50%. A year ago, awareness led at 58%, campaign sales stood at 37%, and conversion at 39%. In one edition of the study, sales rose 14 points and conversion rose 11, while awareness eased.

Even with the directional-comparison caveat applied, the direction is hard to miss: the creator budget is being pulled onto the same scoreboard as performance media. The channel that used to report reach is now expected to report revenue.

The same pattern appears in how marketers rate creator influence across the customer journey. The strongest reported influence sits at conversion and at loyalty, each rated high by 55% of respondents, ahead of consideration (52%), discovery (50%) and repeat purchase (48%). The top of the funnel, where creator marketing built its reputation, is no longer where its practitioners claim it matters most.

For a marketing manager, the practical consequence is uncomfortable and important. A budget that reports against sales needs a chain of measurement that reaches sales. That means tracked links, defined attribution windows, an agreed answer to what counts as a creator-driven purchase, and a reporting method fixed before the campaign launches, not negotiated after the results arrive. The study documents the ambition. It does not supply the plumbing, and its own numbers are a reminder of why: self-reported survey percentages are a description of what marketers believe they measure, not proof that their attribution holds.

The scoreboard shift also rewires selection itself, and in a direction the rest of the study supports. When awareness is the target, the rational buy is reach per dollar, which pushes budgets toward the biggest names. When sales and conversion are the target, the rational buy is audience intent: creators whose followers trust their recommendations on a specific kind of purchase, in a specific category, at a specific moment. That favors smaller, closer, more specialized voices over reach at scale, which is exactly the direction the study's roster figures point, and it changes what a marketing manager should ask to see before approving a creator: not how many people might see this, but what happened the last time this creator's audience was asked to buy something.

There is a warning worth attaching to that shift. Metrics that become targets get managed. Once creator renewal and creator pay ride on attributed sales, every actor in the chain learns to produce attributed sales: last-click credit inflates, discount codes harvest demand that would have arrived anyway, and the cleanest-looking numbers describe the most aggressive crediting rules rather than the most effective creators. A brand that upgrades its scoreboard without upgrading its skepticism about the scoreboard has changed the decoration, not the plumbing.

When the scoreboard changes, the quality of the keeping matters more than the enthusiasm of the players. That is the part of the infrastructure era that does not fit in a press release headline.

What always-on actually commits you to

The study's headline finding still deserves its due, because operating model follows from it. If 91% of surveyed brands run creator programs with an always-on component, then influencer selection has stopped being an event and become a standing process.

An always-on program does not get to make the selection decision once per campaign, with a brief, a deadline and a clean slate. It makes the decision continuously: new creators enter the roster, existing ones are re-evaluated as their audiences drift, prices move, and last quarter's fit is this quarter's open question. The organizational spread the study reports, creator programs touching marketing, social, media and brand teams at once, with affiliate involvement up 7 points to 36% and PR up 15 points to 30%, means the decision also has to survive more hands and more definitions of success than it used to.

This blog has argued before that the operational load of repeated verification is where always-on programs strain. The study adds a measurement of that strain from the buyer's side: 90% of marketers say they want a less complex creator marketing technology stack, and 45% still rely on manual processes, including spreadsheets and email, to run a function they simultaneously describe as infrastructure. The gap between those two numbers is the honest core of the report. The industry renamed the channel infrastructure before it finished building the pipes.

There is also a selection implication hiding in the scale figure. When the average brand works with many more creators than it did a year ago, the unit of analysis changes. Nobody meaningfully reviews a roster of that size creator by creator, campaign by campaign, by hand. The review has to be systematic, or it quietly stops happening. A process that exists on paper and not in practice is worse than an honest absence of process, because it manufactures confidence without evidence.

The economics of cadence cut the same way. In a campaign world, the cost of checking a creator is paid once per launch and amortized against a visible budget line. In an always-on world the cost is continuous and mostly invisible: every active creator is a recurring micro-decision to keep paying, keep briefing and keep publishing, made so often that nobody experiences it as a decision at all. Programs that thrive under that load treat the roster as a portfolio with pruning rules, not only acquisition rules. They define in advance what keeps a creator on the roster and what takes one off, because a review that depends on someone remembering to do it will be remembered less and less often as the roster grows.

The quiet rise of the certified expert

One of the study's quieter movements says a great deal about where brand preference is heading. Lifestyle creators still lead in stated importance at 55%, but category-niche specialists are close behind at 50%, and certified experts, chefs, dietitians and trainers in the study's examples, stand at 49%. Certified experts posted the largest gain of any group, up 18 percentage points in a year.

Read alongside the trust findings, the pattern is coherent. Brands are not only buying reach; they are buying credibility that survives scrutiny. A credentialed creator brings an audience that expects substance, and content that can carry claims a generalist cannot.

The caveat writes itself, and it is one this publication has explored from another angle: a credential is a claim, not a verification. As certified expertise becomes a marketing asset, the incentive to wear the costume grows with it. The study documents demand for experts; it does not examine how brands confirm that the expertise is real, current and relevant to the product. That confirmation is the buyer's job, and the study's own trust data points at exactly the same job from the consumer's side.

Trust is the asset, and the study lists its ingredients

For the second consecutive year, marketers in the study rank creators as the most trusted source of product recommendations: 44% select creators, ahead of social media ads at 34% and celebrities at 22%, and the release notes the gap is widening year over year.

A ranking like that is easy to quote and hard to use. The more operational finding is the study's list of what makes a creator recommendation feel trusted: clear and useful product information (62%), transparency about paid relationships (60%), genuine category expertise (54%) and honest comparisons (51%). Marketers also say creators help consumers feel confident in their decisions (65%), demonstrate how products work in real life (61%) and simplify complex choices (57%).

Notice what those ingredients have in common. Every one of them is checkable before money moves. A brand can review a creator's past content for the quality of its product information, for how it handles paid relationships, for demonstrated expertise and for whether its comparisons are honest. None of that requires waiting for campaign results. The trust that the study calls the system's most valuable asset is not a vibe to be hoped for; it is a set of observable behaviors to be verified at the point of selection.

That reframing matters for how the trust ranking should be used. The 44% figure describes an average across a vast population of creators. It says nothing about the specific creator in front of you. Trust is not inherited from the channel; it is earned by the individual, and it can be assessed individual by individual if someone does the work of looking.

Where the study argues with itself

The most instructive tension in the 2026 release sits between its technology appetite and its operational reality.

On one side: 95% of marketers rate AI as a must-have or important when evaluating creator marketing platforms, looking to it for personalization, performance forecasting, strategic planning and content reuse. For the first time, 41% say they use creator content to influence how products appear in AI-driven search experiences, a use case that did not exist in previous editions. On the other side: the 90% who want a simpler stack and the 45% still running on spreadsheets and email.

A vendor reading resolves the tension in the vendor's favor: the answer to manual complexity is a smarter platform, ideally an agentic one, ideally the one announced in the same press release. A buyer's reading should be more careful. AI features in creator platforms are exactly the kind of claim that deserves the same scrutiny this blog applies to any scoring system: what does the recommendation rest on, can it be explained, and does it survive contact with an experienced marketer's judgment? The study shows demand for those answers. It does not, and cannot, show that any particular platform supplies them.

The AI-search finding deserves the same discipline. That 41% of surveyed brands already use creator content for AI-driven search visibility tells you the practice is spreading. It does not tell you it works, how it is measured, or what it is worth. New channels arrive with enthusiasm first and instrumentation later. The brands that will know whether this one pays are the ones that defined the metric before they spent.

Where Flonci fits

FLONCI's premise lines up with the part of this study that will still matter after the headline is forgotten. If creator budgets now report against sales and conversion, then the selection decision has to be made on a basis that can carry that reporting. If programs are always-on, then verification cannot be a once-a-campaign project; it has to be continuous, systematic and fast enough to keep up with a roster that never stops moving. And if trust is built from observable ingredients, product information quality, transparency, expertise, honest comparisons, then those ingredients can be checked in advance rather than assumed.

That is the decision point Flonci was built around: a layer that sits at the moment an influencer choice turns into a budget commitment, and insists that the choice arrive with a checked basis. Not a promise of outcomes, because no honest system can promise those, and this study does not change that. A basis: fit to the goal, evidence about the creator's actual audience and past commercial work, and reasoning the marketing manager can examine and defend.

The study's own tensions make the same argument from the market's side. Ninety percent of its respondents want simpler operations. Nearly half still work manually. The distance between those facts is where verification either happens systematically or quietly stops happening. Closing it is the work.

Conclusion

The fourth LTK-Northwestern study is a useful document less for its declaration that creator marketing became infrastructure than for the evidence it carries about what that declaration demands. Sales and conversion now share the top of the metric list with awareness. Certified expertise is gaining on lifestyle reach. Trust has identifiable, checkable ingredients. And the operational reality, complex stacks, manual processes, continuous rosters, lags the ambition by a visible margin.

The study is also, transparently, a vendor instrument: 209 self-reported senior marketers at brands already committed to the channel, with year-over-year comparisons its authors label directional. Read that way, it is still one of the better annual mirrors the industry has. The brands that will benefit most from the infrastructure era are not the ones that adopt the vocabulary first. They are the ones whose measurement, vetting and selection discipline can actually carry the weight the new scoreboard puts on them.

Infrastructure is judged by what runs on it. The next time a creator budget is approved, the approval should be able to show its plumbing.

Research notes and sources

This article is based on "LTK and Northwestern University Release Fourth Annual Creator Marketing Study: Creator Marketing Enters Its Infrastructure Era," published October 21, 2026 via Business Wire (businesswire.com), with the study landing page at company.shopltk.com. All figures cited in this article come from that release.

Method and limitations, as stated in the release: the research was conducted by LTK in collaboration with Northwestern University in June 2026, surveying 209 C-suite marketers and senior decision-makers at brands that invest in creator marketing, the majority at organizations with more than $100 million in revenue. The release notes that because the 2026 survey updated its sample and questionnaire, comparisons with prior years should be considered directional. Readers should also weigh that LTK sells the creator marketing infrastructure the study discusses, and that all figures are self-reported by respondents rather than measured from campaign data. Nothing in this article is a verified FLONCI client result or a commitment to an outcome. The cover image is an original AI-generated illustration, not a product screenshot or a chart of the study's data.

More Followers Do Not Necessarily Produce More ROI

Original AI-generated illustration for Flonci. Conceptual artwork, not a product screenshot.

ROI data from Israel and campaign examples from Samsung, L'Oreal and Virgin Voyages point to a new pattern: less betting on one big name, more building a portfolio of creators, measuring, and reinforcing what proves itself.

For years, one of the simplest ways to choose an influencer was to look at the biggest number on the profile: how many followers he has.

More followers seemingly meant more exposure. More exposure justified a higher price. And so, the more important the campaign, the stronger the tendency to choose the bigger name.

The data is starting to tell a more complicated story.

The Israeli benchmark that frames the question

In Humanz's Israeli benchmark for 2026, based according to the company on hundreds of organic campaigns measured in its system, the average ROI of Micro influencers stood at about x5.16. For A-list influencers it stood at about x5.2. Nearly the same number.

But moving from the average to the median opens a gap: the median among Micro influencers stood at x4.8, compared with x4.1 among A-list.

That does not say a small influencer is better than a big star. It does place a question mark over one of the market's basic assumptions: audience size alone does not predict the return on the investment.

And the conduct of some of the world's largest brands shows they are already working accordingly.

Samsung: instead of choosing one tier, test them all

Samsung did something especially interesting in a TikTok campaign in Germany for the launch of Galaxy products.

Instead of building the move around a few big names, the company worked with 100 creators, from Micro to Top-tier, and produced 106 content assets. Over four weeks it rotated the ads at a high frequency, every three to four days.

The results illustrated how much a tier is only a starting point. One Micro creator reached 26% engagement, while another creator in the same layer reached only 1.19% (the low-end figure as cited in the original Hebrew article; it does not appear in the archived case-study summary). Average engagement in the Mid-tier was 6.2%, and in Macro about 8.1%. The Top-tier showed less variance and therefore more predictable results, but demanded a higher investment.

Samsung did not discover that Micro creators "win". It discovered something more useful: within the same group of influencers, the gaps in performance can be enormous.

So the way to benefit from the advantage of smaller creators is not to pick one Micro and hope for the best. You work with a sufficient number of creators, measure, and identify the ones who truly deliver.

According to TikTok for Business, the campaign led to a 125% increase in average watch time, a 141% increase in CTR and a 53% decrease in CPC on creator-based consideration assets. These are the platform's case-study figures, so they are best read as a business example and not as independent market research.

L'Oreal: the creator is only the first stage

L'Oreal also chose not to pit Micro against Macro.

In the #BeautyTok campaign it used both groups together. The creators received products and produced content in their natural style. Then came the important stage: L'Oreal took the creative and ran different media layers on it, including TopView, In-Feed and Spark Ads.

In other words, choosing the influencer was not the end of the decision. It was the beginning of a system: creator, content, measurement, amplification.

According to TikTok for Business, the campaign was also measured against business results, including a 43% increase in e-commerce sales in Peru and 92% in Ecuador.

Here too, the question is not "which tier won?". What is interesting is that the brand built a mix of creators, formats and media, rather than assuming the influencer's size alone would determine the campaign's success.

Virgin Voyages: 1,103 creators instead of a bet on a few stars

Virgin Voyages took the idea one step further.

In a move published in 2026, 1,103 creators took part, combining Micro creators, larger talent and community activity. The move produced more than 22.7 million video views and more than 3,200 posts around the campaign.

But the most interesting detail is not the number of creators. Virgin received 180 days of content usage rights, so it could identify which videos performed better and then keep promoting and distributing exactly those, through paid media.

The same structure returns: do not pick a winner in advance. Create variety, measure, identify winners, and then invest more in them.

Three campaigns, one pattern

Samsung, L'Oreal and Virgin Voyages are very different from one another. Technology, cosmetics and travel. The campaign goals and the markets differ as well. But the same logic appears in all three.

Do not rely on a single tier. Micro, Mid, Macro and A-list fill different roles. Do not assume the follower count predicts performance. Even within the same layer of creators, the gaps can be extreme. Do not treat the creator's post as the finish line. The content becomes an asset that can be measured, compared and amplified. Move budget to what proves itself. The budget turns from something allocated in advance into something managed throughout the campaign.

And this connects directly to the Israeli figure we started with. When the average ROI of Micro and of A-list is nearly identical, x5.16 versus x5.2, it is hard to justify a choice based on audience size alone.

The marketing manager's question changes

Less: who is the biggest influencer we can buy? More: which creator is expected to produce the best value for our price, our goal and our audience?

And perhaps even more important: how do we build a campaign that lets us discover that before we allocate most of the budget?

This is no longer just influencer selection. It is investment management in influencers.

Research notes and sources

This article is an English edition of "יותר עוקבים לא בהכרח מייצרים יותר החזר השקעה" by Amit Magen, first published in Hebrew on October 7, 2026 and preserved in the FLONCI archive at flonci.co.il. The original archive notice applies: the content reflects its original publication date; data, prices, examples and tool descriptions may have changed, and nothing in it is a verified FLONCI client result or a commitment to an outcome.

Source framing for the figures. The Israeli benchmark figures were checked on October 21, 2026 against Humanz's published 2026 benchmark at benchmark26.humanz.co.il, which describes the report as based on hundreds of organic campaigns measured in Humanz's system, without paid promotion; the tier averages (Micro x5.16, A-List x5.2) and medians (Micro x4.8, A-List x4.1) match that page. They remain Humanz's own measurements, not an independent audit. The Samsung, L'Oreal and Virgin Voyages figures are platform and press case-study numbers, not independent market research. The L'Oreal #BeautyTok figures, including the sales increases of 43% in Peru and 92% in Ecuador, match TikTok for Business's case study at ads.tiktok.com. The Virgin Voyages figures (1,103 creators, 22.7 million video views, over 3,200 posts, 180 days of content usage rights) match TikTok for Business's Creator Voyage case study at ads.tiktok.com; press coverage of the same cruise by Skift and Ad Age confirms the event with its own counts. The Samsung Germany campaign figures (100 creators, 106 content assets, the four-week run, the tier engagement bands, and the 125% watch-time, 141% CTR and 53% CPC changes) follow TikTok for Business's case study, corroborated by a LinkedIn post from TikTok's DACH general manager and an archived reading of the case study; the case study's live page is no longer available, and the single low-end Micro engagement figure of 1.19% is cited from the original Hebrew article, as noted in the text. The cover image is an original AI-generated illustration, not a product screenshot.

A Viral Reel Is Not a Campaign Forecast: What to Check in an Influencer's Commercial Performance

Original AI-generated illustration for Flonci. Conceptual artwork, not a product screenshot.

An influencer sends over a video with hundreds of thousands of views. The content is funny, the comments are flowing, and the brand manager is already picturing the product inside the next reel. Before approving the partnership, one more question is worth asking.

That question is simple: are these the results of his commercial content, or of an exceptional organic video?

The difference matters to the decision. A video that entertains an audience and a video that presents a product do not have to do the same work. Even when the same creator appears in both, the brief, the message and the action asked of the viewer can be entirely different.

This article offers a working method for checking the proposal in front of you. It is not market research, it presents no Flonci performance data, and it promises no result in your next campaign.

Compare content that does similar work

Before drawing conclusions from the most successful reel, ask to see several relevant commercial collaborations. Prefer content from the same platform, in a similar format and from a recent period, with products and an audience that can reasonably be compared to your campaign.

Not every comparison will be perfect. Brands do not always get access to all the data, and sometimes only a small number of similar campaigns exist. That limitation should appear in the decision itself. Limited information is not necessarily a reason to drop the creator; it is a reason not to present the estimate as if it were certain.

Next to each piece of content, note what its goal was. Exposure for a new brand, reaching a product page, and collecting leads are different goals. A video can succeed at one goal and say nothing about another.

Separate the creator from the media

When you are shown a view count, find out whether the content received paid promotion. If it did, ask which figures refer to organic distribution and which to paid media, to the extent the separation is available.

There is nothing wrong with paid promotion. The problem begins when a decision about the creator's distribution strength rests on a number that also includes a media budget not shown in the proposal.

If the sources cannot be separated, write that down. You can still learn from the content and the response to it, but it is not right to use the combined number as independent proof of what the creator will bring without promotion.

Read the comments in context

The comment count is a starting point. After it comes a simple check: what are people reacting to?

Replies to a joke, questions about the product, requests for a link and conversations about price tell different stories. There is no need to classify every comment as purchase intent. It is worth checking whether the conversation that formed is relevant to what you are trying to achieve.

Here too, it is important not to overreach. A question about the product is not an order, and a like is not a sale. These are signals that can help you understand the response to the content, not a replacement for traffic or purchase data when those are needed for measurement.

Do not let one video set the estimate

Ask to see the broader picture, including commercial content that did not perform especially well. If several videos are shown, examine the range between them and the typical result, rather than building the entire decision on the peak.

You can use the median alongside the range to describe the sample you received. But when the sample is small, or selected by the creator, even a median is not a reliable forecast on its own. It is a way to summarize the data shown, not a promise about your campaign's performance.

It is also worth asking what differed between the videos: the product, the opening, the message, the posting time or the call to action. It will not always be possible to know what caused the gap. The very act of distinguishing between what is known and what is assumed improves the quality of the conversation.

Connect the check to your brief

In the end, the goal is not to give the influencer a general score. The goal is to understand whether there is a reasonable basis for the specific collaboration you are planning.

Before approving, write down: What do we want the viewer to do? Which similar commercial content did we examine? What is known about the audience exposed to it? Do the figures include paid media? How will we measure the result on our side? What information is missing, and how does it affect the scope of the commitment?

If the goal is traffic to the site, define a tracked link and a reporting method in advance. If the goal is sales, define what counts as an attributed purchase and what the attribution limits are. If the goal is awareness, choose a suitable metric instead of promising sales that were never defined as the campaign's goal.

Start at a scope that matches the level of certainty

When there is a good fit but little information, a limited initial activity with orderly measurement is worth considering. Decide in advance what you want to learn, which finding would justify continuing, and which usage rights are needed already at the first stage.

A small test does not guarantee that the next stage will succeed. It gives the brand a chance to learn about the connection between the creator, the content and the product before a broader commitment.

The viral reel can still be a reason to pick up the phone. It simply should not be the only document the budget is approved on.

Before choosing an influencer, check what stands behind the performance you are being shown, and how it relates to the work you are asking him to do.

Research notes and sources

This article is an English edition of "ריל ויראלי הוא לא תחזית לקמפיין: מה לבדוק בביצועים מסחריים של משפיען" by Amit Magen, first published in Hebrew on October 9, 2026 and preserved in the FLONCI archive at flonci.co.il. The original archive notice applies: the content reflects its original publication date; data, prices, examples and tool descriptions may have changed, and nothing in it is a verified FLONCI client result or a commitment to an outcome. The cover image is an original AI-generated illustration, not a product screenshot.

FLONCI: The Israeli Development Connecting AI to Influencer Budget Approval

Original AI-generated illustration for Flonci. Conceptual artwork, not a product screenshot.

The launch is tomorrow. The influencer decision is due today. A look at how FLONCI connects data, brand fit and judgment at the moment a marketing budget is committed.

A fashion company runs dozens of campaigns a year around capsule collections. Each launch replaces the products, the story and the audience it wants to reach. One capsule speaks to young women chasing the next trend. Another is built for women looking for workwear. For every launch, the marketing team has to match the move with the influencers who can connect the product to the people who might want it.

Now one of those collections goes live on the site tomorrow morning. The photos are ready. The inventory has arrived. The budget is approved. What remains is choosing the influencers and starting to coordinate the activity, and the decision has to be made today, because every delay in choosing shrinks the time left for production, approvals and publishing.

This is the challenge FLONCI was built around. FLONCI is a new Israeli development in marketing technology: a decision layer based on artificial intelligence and data for influencer selection, with a clear premise. Turn information into a practical choice, shorten the path from brief to approval, and let marketing managers move forward with confidence.

This article is an expanded English edition of a piece first published in Hebrew on the Flonci insights blog. It walks through the same scenario in more depth: what the decision actually involves, why it gets heavy, and what a protection layer at the point of budget commitment is meant to do.

A list of names is only the beginning of the work

At the moment a launch needs influencers, a list of names is where the work starts, not where it ends. Behind every name sit the questions that decide whether the budget is well spent. Who is this product actually for? How relevant is this creator's audience to that buyer? What justifies the cost of this particular creator? And which mix of creators, together, serves the launch within the budget that exists?

Each of those questions has a real answer, and the answers differ from launch to launch. Audience overlap, age distribution, content style, past brand work, engagement quality and pricing all move independently. A creator who looks right on a follower count can be wrong on the audience behind it. A creator who looks expensive can be cheap relative to the buyers they actually reach.

The workwear capsule makes this concrete. An influencer with a large, highly engaged audience looks like a natural choice at first glance. But if most of her audience is very young and shops a different kind of fashion, the fit is weaker than the numbers suggest. A creator with a smaller audience, one that matches the working women the capsule is built for, may be the more relevant choice for the task. The decision has to rest on data and on the campaign's goal, not on the size of the following.

None of this is a new insight for experienced marketing managers. It is what good influencer selection has always required. The problem is not knowing what to check. The problem is the time it takes to check it, launch after launch, when the calendar gives you a day.

The mix question: one budget, several creators

The hardest part of the decision is rarely the evaluation of any single creator. It is the composition. A launch budget has to be divided across creators whose audiences, formats and prices differ, and the best individual choice is not always part of the best mix. Two strong creators whose audiences overlap heavily can deliver less together than a smaller pairing with complementary reach. A single expensive name can absorb the budget that three mid-sized creators would have turned into broader coverage.

For the workwear capsule, the mix question sounds like this: does the launch need one authoritative voice that working women already trust, or several smaller voices across the segments the collection covers? Should part of the budget go to a creator known for styling advice, whose content keeps working for weeks, alongside creators whose strength is immediate launch-day reach? These are portfolio questions, and they cannot be answered by looking at any profile on its own.

This is also where the data work multiplies. Comparing mixes means evaluating every candidate against the goal and then against each other, inside a fixed total. Done by hand, under a one-day deadline, the comparison usually collapses into a shortlist built on familiarity. The launch still happens. It just carries an unexamined mix behind it.

The operational weight of repeated verification

A company that manages a steady sequence of launches does not make this decision once. It makes the decision every few weeks, under the same time pressure, with the same rigor required each time. Checking accumulates into an operational load: hours of profile review, audience analysis, price comparison and internal alignment, repeated across every capsule, every season, every market.

Those hours have an opportunity cost that rarely gets counted. Every hour saved on verification is an hour that can go to coordinating with the creators themselves, sharpening the message, and preparing the content the launch depends on. In a company running many launches, shortening the path to a decision does more than save time. It expands the organization's ability to respond to opportunities: a trend that breaks on Tuesday can still make it into Thursday's plan.

Speed alone is not the goal, and this is where the framing matters. A fast decision with a weak basis is just a faster mistake. The value lives in the combination: a decision made in time, with its fit to the goal and the budget actually checked. That combination is what a decision layer is for.

A protection layer at the point of budget commitment

One way to understand FLONCI's approach borrows an image from cybersecurity. Security teams talk about a protection layer that sits at the point where exposure is created, not after the damage is done. The marketing version of that point is the moment an influencer choice turns into a budget commitment. Before it, a mismatch is a cheap correction. After it, the same mismatch is spend that cannot be called back.

The purpose of a protection layer at that point is not to make the decision for the marketing manager. It is to surface a misfit early, while it is still a finding rather than a loss, and to give the investment decision a clear basis before the money moves. Audience does not match the product. Pricing does not match the expected contribution. The mix does not cover the segments the launch needs. Each of these is cheaper to learn before approval than after.

For the fashion company in the scenario, the practical meaning is the ability to examine every capsule according to its own characteristics. The trend-driven capsule for young women and the workwear capsule for professionals are different products for different buyers, and they deserve different creator mixes, even inside the same season and the same budget line.

A familiar creator still earns a fresh check

A subtle part of that idea deserves its own attention. A creator who fit the previous launch can be examined again against the product, the audience and the goal of the next one. Fit does not transfer automatically between capsules, because the product changed, the audience for it changed, and the creator's own audience keeps moving.

This is not an argument against working with creators the company knows. Accumulated experience and familiarity have real value: a known creator is faster to brief, easier to coordinate, and more predictable in production. The argument is for keeping that experience alongside a focused check of the current need. The previous fit is evidence. It is not an exemption.

In practice this is the difference between a roster and a decision. A roster says who the company likes to work with. A decision says who serves this launch, for this audience, at this price. A protection layer exists to keep the second question alive even when the first answer is comfortable.

Three questions to demand before approving an influencer

The same discipline is available to marketing managers directly, whatever tools they use. It comes down to demanding that every influencer choice arrive with an answer to three simple questions.

First: why does this influencer fit the target audience? A useful answer names the audience the launch is for and shows the overlap with the creator's actual followers, not their follower count. Second: what is this influencer supposed to contribute to the campaign's goal? A useful answer connects the creator to the specific job of the campaign, whether that is reach, credibility in a niche, content production or conversion. Third: what is the estimate that the investment in this influencer is justified based on? A useful answer makes the reasoning inspectable: the pricing, the expected contribution, and the comparison to the alternatives that were considered.

Clear answers do two things at once. They let alternatives be compared on the same footing, and they let the decision be explained inside the organization. When the finance partner or the CEO asks why this creator and not another, the answer already exists, because it was part of the approval.

The collection launching tomorrow shows why this matters. It needs a decision that can be made now, and also one that can be explained afterward. Those two requirements are not in tension. They are the same requirement: a basis clear enough to act on.

The decision has to survive internal review

Influencer budgets do not get approved in a vacuum. A marketing manager's choice has to hold up in front of a finance partner asking about expected return, a CEO asking why this creator and not the one a competitor uses, and sometimes a board asking what the influencer line actually bought last quarter. A decision that cannot be explained tends to get re-litigated, and re-litigation costs the same hours the original verification did.

The three questions double as the explanation. When every approved choice arrives with its fit, its expected contribution and the basis for its price, the internal conversation moves from opinion to evidence. The finance partner can examine the same reasoning the marketing team used. The CEO's question has a documented answer. The next quarterly review has a record of why each commitment was made, which makes the results review honest: it can compare what was expected with what happened.

This organizational layer is easy to underestimate because it produces nothing visible. No post, no story, no launch asset. What it produces is alignment: a decision the whole company can stand behind, made once, at the speed the launch calendar demands.

How to test the business value of AI

For technology and innovation leaders, a decision layer like this is also a practical test of what AI is worth in a marketing organization. The test does not require faith. It requires measurement, and the measures are concrete.

The first measure is time: how long does it take to get from a brief to an approved choice, before and after. The second is manual work: how much of the verification load moved off the team's calendar. The third is usefulness: how often does the recommendation survive contact with an experienced marketing manager's judgment, and how often is it rightly overridden. A layer that is always overridden is noise. A layer that is never questioned is a risk. The healthy pattern is a recommendation the team engages with and usually builds on.

The longer-run measure arrives with the campaigns themselves. Results make it possible to examine the quality of the choices and learn from them: which fits predicted well, which mismatches slipped through, and what the next launch should check earlier. A decision layer that cannot be audited against outcomes is a black box, and marketing budgets should not fund black boxes.

What launch week looks like with the checking done

Return to the fashion company on the morning the decision is due. In the unassisted version of the day, the team splits its hours between gathering audience data, reconciling prices, assembling a shortlist and aligning internally, and the creative coordination waits for all of it. The creators who do get chosen hear from the brand late in the day, briefs go out compressed, and the content window that was already tight gets tighter.

In the version where the verification layer has done its work before the day begins, the same team starts from an examined shortlist with the three questions answered per candidate. The morning goes to the work only people can do: talking to the chosen creators, shaping the briefs, fitting the message to the collection's story. The launch does not just get a faster decision. It gets a better-prepared campaign, because the hours moved from checking to building.

Multiplied across a year of capsules, that reallocation is the real return on the layer. Not a single saved afternoon, but a standing change in where the team's time goes: less gathering, more deciding; less deciding, more creating.

What FLONCI is built to do

The promise FLONCI is working to deliver follows directly from the scenario: a faster choice, with a clearer basis for the investment. A protection layer that accompanies marketing managers at the moment information turns into a budget commitment, and lets them set out with a decision they understand and can explain.

The premise is deliberately conservative about where the value sits. The creator's value does not change because the tooling does, and no layer replaces the marketing manager's judgment about brand, message and timing. What changes is the cost of verification: the data work behind the three questions arrives faster and more consistently than a team under launch pressure can produce it by hand, and the decision that remains is the human one.

For a company managing many launches, that shift compounds. Each capsule gets examined against its own product, audience and goal. Each approval carries its reasoning with it. And the hours that used to go to gathering the basis for a decision go back to the work the decision was meant to enable.

Conclusion: a decision you can explain

The fashion company's collection will go live tomorrow either way. What is decided today is whether the influencer budget behind it was committed with a checked fit, a clear contribution and a stated basis, or on the strength of a familiar name and a short clock. The first kind of decision is not slower. It is just better supported, and support is exactly what a decision layer exists to provide.

That is the development FLONCI represents: connecting AI, data and judgment to the influencer budget approval process, at the point where the connection matters most. Know before you spend. The next influencer you are about to approve is worth one good check.

Research notes and sources

This article is an expanded English edition of "FLONCI: הפיתוח הישראלי שמחבר AI לתהליך אישור תקציבי משפיענים" by Amit Magen, first published in Hebrew on October 21, 2026 on the Flonci insights blog, flonci.co.il. Product direction and company descriptions reflect the company's own published materials at flonci.com and flonci.co.il.

Two disclosures carry over from the original. The fashion company in the article is a fictional illustrative scenario, not a customer story. The cover image is an original AI-generated illustration, not a product screenshot.

Before Creators Move to TV, Check What the Survey Actually Asked

Original AI-generated illustration for Flonci. Conceptual artwork, not a chart of the research results.

A new survey says creators on TV move purchase intent, especially for Black and Hispanic audiences. The evidence deserves a careful read before any budget moves.

A creator with a loyal following appears in a streaming series. A brand's product sits inside the scene. The pitch writes itself: the trust the creator built in the feed now works on the biggest screen in the house, at television scale.

The Video Advertising Bureau, a trade body for TV and streaming companies, published a report on October 15, 2026 that appears to confirm the pitch. Its headline finding: 53% of Black respondents who follow creators say they would be more likely to buy a brand advertised in, or integrated into, a TV or streaming show featuring a creator they follow. Hispanic and Black respondents follow creators at 71% and 64%, well above the 55% average.

The numbers travel well in a sales deck. They also deserve the questions any brand should ask before moving creator budget into premium video: who exactly was asked, what were they asked, how large are the groups behind the percentages, and what would the same people have said about a plain social video?

The answers, documented in unusual detail by PPC Land's review of the 22-page report, make this a useful case study in reading creator research before acting on it.

A new release built on data that is not new

The report is titled "Creating Credibility: How Premium Video Enhances the Connection Between Digital Creators and Diverse Audiences." It was produced with Hub Entertainment Research and released in New York on October 15.

None of its data is new. VAB calls it a companion piece to "Expanding the Creator-Verse," published in August. Both draw on the same source: a June 2026 online survey of 1,600 respondents aged 16 to 74. The August version reported the results by age group. The October version reports the identical totals re-cut by ethnicity: Hispanic, Black, and non-Hispanic white.

This matters for how the release should be read. It is not a fresh October study. It is a second look at a June survey, published four months later, structured to support the same conclusion from a new angle. Re-analysis is legitimate, but a reader should know the headline "new research" is new packaging.

It also matters who is making the argument. VAB's members are, in its own words, premium multiscreen TV providers and distributors. The report concludes that creator marketing money performs well when it flows through those members' platforms. That does not make the data false. It makes independent scrutiny part of the price of reading it.

The headline group is 122 people

Every behavioral figure in the report describes only respondents who follow at least one creator: 889 of the 1,600 surveyed. Within that group, the ethnic breakouts rest on 193 Hispanic respondents, 122 non-Hispanic Black respondents, and 505 non-Hispanic white respondents. The three subgroups sum to 820, leaving 69 creator-followers who appear in the totals but in none of the breakouts.

The 53% purchase-intent headline therefore describes about 65 people. VAB publishes no margin of error. PPC Land's own calculation, using the standard formula for a simple random sample, gives roughly plus or minus nine percentage points for a base of 122 and seven for a base of 193, and online panels are not random samples, so the true uncertainty is likely wider.

With intervals that size, the 53% Black figure and the 45% Hispanic figure overlap substantially. The report's strongest single number is also its least stable. What carries more weight is that the direction of the differences repeats across all five funnel stages; a consistent pattern across noisy measurements says more than any one precise-looking percentage.

There are also signs of hasty production. A footnote about follow rates describes the wrong base. Summary charts carry the subtitle "Top 5 Selections by Age Demo" on charts that are broken down by ethnicity, apparently copied from the August deck. An appendix chart cites the wording of the wrong survey question. None of these change the data, but they signal how much review the document received.

The question measures what people say they would do

The core of the report is one question asked five times: if a creator you follow appeared on a TV or streaming show, how likely would their involvement make you take each action, from paying attention to the brand through buying it? Answers used a five-point scale, and VAB reports the top two boxes against the bottom two as a ratio.

Three properties of that design shape everything downstream. First, the scenario is hypothetical. Respondents described what they think they would do, not what they did. Stated intent systematically overstates observed behavior, and the survey contains no behavioral check.

Second, the middle of the scale disappears in the reporting. For purchase among all creator-followers, 43% said more likely and 18% said less likely, which means 39% said the creator's presence would make no difference. The 2.4-to-1 ratio sounds like a wave of influence; the same data says the single largest group is indifferent.

Third, the ratio format hides the negatives. Hispanic respondents follow creators more than any other group at 71%, yet they also gave the highest "less likely" answers of any group, 21% to 22% on attention, consideration, and purchase. The release quotes the 45% Hispanic purchase figure and leaves out the 21% negative. In this data, the audience most devoted to creators is also the audience most likely to say a creator's TV appearance would put them off a brand.

There is no control condition

The deepest limitation is what the survey never asked. Nobody was asked how they would react to the same creator promoting the same brand in an ordinary social video. The question was written to test exactly the scenario VAB's members sell: creators inside premium video.

Without that comparison, the survey cannot separate what the creator adds from what the TV setting adds. If 43% of followers say they would be more likely to buy after seeing a creator on TV, and a similar share would say the same about a creator's unaided social post, the premium video premium is an assumption, not a finding. The survey is structurally incapable of detecting it either way.

This is the single most important question a brand can carry out of the report, because it defines the research that would actually justify the spend: a test that compares the same creator, the same brand, and the same audience across the two environments.

What the survey does support

Read with those limits, the report still contains usable evidence. The first is about how audiences categorize creators. Asked which descriptions fit online creators today, every group put "entertainers" first: 56% of Hispanic respondents, 51% of Black respondents, 55% of non-Hispanic white respondents. Trusted experts scored 16% to 18% across groups.

The second concerns what builds credibility. Non-Hispanic white respondents put expertise and knowledge first at 43%, and Black respondents also put expertise first at 38%. Hispanic respondents put an authentic, active community presence first at 40%, with higher production quality second at 37% and expertise third at 31%. Black respondents named partnerships with brands as a credibility signal at 28%, seven points above the other groups, which is the report's clearest demographic difference.

The third cuts against a familiar pitch. Only 15% of Hispanic respondents described creators as feeling like a friend or neighbor, compared with 24% in both other groups. The parasocial-friend framing common in advertiser decks is weakest in the group that follows creators most. Audiences appear to value creators as performers and credible sources, not as pretend friendships.

And the fourth is awkward for the report's own thesis. Asked what makes a creator feel established, appearing in TV or streaming programs scored 22% to 24% across groups, below production quality, authenticity, expertise, and consistent posting in every group. The survey's own data says audiences value the qualities TV can support, not the TV appearance itself.

The funnel multiples, read group by group

The full funnel results are worth laying out, because the group pattern is more informative than any headline. On paying attention to the brand, Black respondents led at 59% more likely versus 20% less likely, with Hispanic respondents at 49% versus 22%, non-Hispanic white respondents at 48% versus 14%, and all followers at 50% versus 17%. On consideration the ordering was similar: 57%, 45%, 46%, and 48% more likely respectively.

Two features of that table deserve a second look. The first is that the largest ratios do not belong to the group the release features. Non-Hispanic white respondents produced the strongest ratios on attention, consideration, and search, because their "less likely" answers were the smallest. The release leads with the groups where the positive answers are largest, not where the balance of opinion is most favorable. Both are legitimate readings; a buyer should know which one is being presented and why.

The second is the consistency. Across all five stages and all three groups, more people said the creator's presence would help than hurt. Whatever the caveats about scale and sincerity, the directional finding is stable: creator-followers generally expect a creator's TV appearance to help a brand rather than harm it. A directional finding like that justifies a test. It does not justify a forecast.

Entertainment context is not endorsement

The entertainers-first finding has creative implications that are easy to miss. If audiences primarily experience creators as entertainers, then a creator's appearance inside an entertainment product is a continuation of their role, not a departure from it. The integration borrows the mood of the show and the creator's presence, which is exactly why it feels natural, and exactly why it should not be confused with a recommendation.

Compare that with how the same audiences treat expertise. Expertise ranked first or near the top as a credibility builder in every group, yet only 16% to 18% described creators as trusted experts. The credibility audiences grant is real but specific: it attaches to what the creator demonstrably knows and consistently does, not to their presence in a polished environment.

For a brand, this shapes what the integration should ask the creator to do. A placement that trades on the creator's demonstrated knowledge, such as a cooking creator inside a cooking format, aligns the entertainment context with the credibility source. A placement that parks a famous face next to an unrelated product borrows attention without borrowing trust, and the survey's credibility data suggests audiences can tell the difference.

Search is the bridge worth watching

One funnel stage deserves attention from performance marketers. On searching for the brand online, the survey found the widest positive ratios: 52% of all creator-followers more likely versus 13% less likely, and among non-Hispanic white respondents 51% versus 12%, a 4.3-to-1 ratio.

If stated behavior became real behavior, the mechanism is familiar: a viewer sees the creator in the show, does not click anything because television has nothing to click, and later types the brand into a search box. The response to creator-led TV would surface in branded search volume, paid search efficiency, and direct traffic rather than in any social dashboard.

The practical implication is about measurement design, not about buying anything yet. A brand testing creator integrations in premium video should instrument branded search before launch, with a baseline and a control market if possible, because that is where this survey says the signal would appear.

What to ask before the budget moves

The report's underlying direction is plausible. Creators do command unusual audience relationships, and putting trusted faces into long-form environments may well transfer some of that trust. Plausible directions are exactly where disciplined buyers earn their keep, because plausibility is what weak evidence wears to meetings.

A brand considering creator-led premium video can ask for five things. Observed behavior from any comparable past integration, not stated intent. Category-specific results, because a snack and a software subscription do not share a purchase journey. Sample sizes behind every demographic claim. A comparison against the same creator's ordinary social performance. And a pilot structure with a control, so the first spend produces evidence rather than confirmation.

None of these asks is hostile to the idea. They are the difference between buying a hypothesis and buying a result.

A hypothetical budget decision

Consider an illustrative decision, not a finding from the survey. A beverage brand's agency proposes shifting 20% of the creator budget into integrations inside two streaming series that feature food and lifestyle creators. The deck cites the 53% figure, the 71% follow rate, and the 4-to-1 search ratio.

The disciplined review does not reject the idea. It re-scopes it. The first commitment becomes a single-show pilot in two matched markets, with branded search and a brand-lift study as the instruments, sized so that a null result is visible rather than explainable away. The creator's own social channels run comparable content in the same window, giving the brand its missing control condition: the same creator, the same product, different environments.

The 20% shift only happens if the pilot separates the creator effect from the screen effect. If the social content performs the same as the integration, the brand has learned something worth more than the placement: its creator budget was already working, and the premium video premium was packaging.

The point of the example is that the survey's weaknesses do not make the channel wrong. They make it unproven, and unproven channels are bought with experiments, not with annual commitments.

Keep stated intent and observed behavior on separate lines

Stated-intent surveys have a real use. They map expectations, language, and relative appeal, and they can rank options cheaply. What they cannot do is forecast the size of a behavior change, because people are poor reporters of their future selves.

A planning process that respects this keeps the two on separate lines of the same document. Stated intent can justify exploring a channel and shaping a pilot. Only observed behavior can justify scaling one. When a single number is asked to do both jobs, as the 53% headline is, the plan has quietly promoted a map into a measurement.

The same discipline applies inside the organization. The person presenting the research should be able to say, in one sentence, what would have to be true for the number to mislead us. For this survey the sentence is short: the groups are small, the questions are hypothetical, the negatives are underreported, and nobody was asked about the alternative.

Where Flonci fits

For Flonci's readers, this is a decision-before-spend question. A creator's value does not change because the screen does, and the evidence standard should not drop because the pitch improves.

The evaluation is concrete. What did the research actually measure, and how large are the groups behind the persuasive figure? Does the proposal compare the creator's effect across environments, or assume the new one is better? What observed behavior will the first spend produce, and what result would stop the scale-up?

No trade-body survey can answer those questions for a specific brand, category, and creator. The research describes what audiences say. The decision about what they do still belongs to the people making the spend, measured where the behavior actually happens.

Conclusion: a promising direction, an unfinished case

The VAB report documents real patterns: creator-following is higher among Hispanic and Black audiences, expertise and authenticity drive credibility, and creator presence in premium video generates strong stated intent, especially toward search.

It does not document that premium video adds anything to what the creator already delivers, because it never asked. The headline rests on about 65 people's predictions of their own behavior, reported without uncertainty intervals, by an organization whose members sell the inventory being studied.

Treat the channel as a hypothesis worth testing. Instrument branded search, run the comparison the survey skipped, and scale on observed behavior. Know before spend, and be suspicious of any number that arrives already certain of itself.

Research notes and sources

PPC Land, "53% of Black creator followers would buy brands in their TV shows, VAB finds," October 15, 2026. Primary source for this article. An unusually detailed critical review of VAB's 22-page report, including base sizes, question wording, and template errors; the margin-of-error calculations are PPC Land's, not VAB's. ppc.land

Video Advertising Bureau, "Creating Credibility: How Premium Video Enhances the Connection Between Digital Creators and Diverse Audiences," October 15, 2026, with Hub Entertainment Research. The underlying report. June 2026 online survey of 1,600 US respondents aged 16-74; ethnic breakouts based on 193 Hispanic, 122 Black, and 505 non-Hispanic white creator-followers; stated-intent methodology with no control condition. VAB is a trade body for TV and streaming providers. The report PDF is marked for members and qualified marketers; this article relies on PPC Land's documented review rather than a direct read of the deck.