New research on hundreds of thousands of creators suggests that small accounts deserve a portfolio decision, not a simple engagement-rate verdict.
A marketing team is choosing between one recognizable creator and a group of smaller accounts. The large creator has a clear price, an established process, and enough followers to make the presentation look convincing. The smaller creators have attractive engagement rates, uneven results, and a much longer list of conversations to manage.
It is tempting to settle the choice with a familiar rule: smaller creators are more authentic, or bigger creators deliver more scale. Neither rule answers the actual budget question. What matters is the outcome being bought, the price paid for it, and the uncertainty the team can carry.
A new Virev study published on October 5, 2026 gives this debate a more useful starting point. Drawing on public Instagram and TikTok data, the study compares normal and sponsored content, estimates prices from published rate tables, and simulates campaigns with different numbers of creators.
Its headline is striking. In its Instagram simulation, a portfolio of 100 nano creators generated a median total of 969,000 plays, compared with 144,000 for one macro creator. But the important part is the sentence that comes after the headline: these were random draws from historical posts, priced with modeled rates, not a controlled test of two campaigns with equal budgets.
Read carefully, the research is not permission to replace every large creator with 100 small ones. It is a reason to stop treating follower size as a complete purchasing strategy.
What the study actually compared
Virev is a creator-discovery software vendor. Its research uses its own database, so this is commercially produced observational evidence, not an independent census of the creator economy. That distinction belongs alongside the findings, rather than buried after the recommendation.
The Instagram sample contains 391,731 public creator accounts with at least 1,000 followers. The TikTok sample contains 262,498 creators. The underlying posts were dated from August 18, 2025 to September 27, 2026, and the crawler read accounts between August 6 and October 1, 2026. The source sets minimum requirements for measurable and normal posts and removes accounts labeled as businesses, brands, media, or organizations.
The study defines nano creators as accounts with 1,000 to 10,000 followers, micro as 10,000 to 100,000, mid-tier as 100,000 to 500,000, macro as 500,000 to one million, and mega as one million or more. Those are this study's bands. Teams should not assume that another report uses the same definitions.
For campaign comparisons, the researchers repeatedly selected distinct creators within a tier and one of each creator's historical sponsored reels. They added the recorded plays and engagement, assigned prices from published tables, and repeated the process 10,000 times.
This is a simulation built from observed content performance. It describes what combinations of past posts look like under the model. It does not establish what a particular brand's next brief will achieve, whether the audiences are relevant buyers, or whether anyone purchased.
The engagement advantage depends on the denominator
The first useful finding is about measurement. Smaller accounts look much stronger when likes and comments are divided by follower count. In the study's Instagram tier summaries, nano creators' usual reels recorded likes and comments equal to 2.1% of followers, compared with 0.5% for mega creators.
That seems to support the familiar claim that smaller creators have more engaged audiences. But the picture changes when the denominator becomes recorded plays. The reported likes-and-comments-per-view figures were 3.3% for nano creators and 3.4% for mega creators.
These are ratios of platform counts, not the probability that a unique person responds. A play is not necessarily a distinct viewer, and a viewer can leave more than one comment. Still, the comparison makes an important planning point: a higher engagement rate per follower does not automatically mean a stronger reaction among the people who actually saw the content.
The source also reports a much higher ratio of normal reel plays to followers for smaller accounts. But that ratio is not the percentage of followers reached. Reels can travel beyond the follower base, and plays can repeat. Calling it follower penetration would claim more than the public counts can show.
A better creator assessment keeps these measures separate. Follower count describes account size. Typical plays describe observed distribution. Likes and comments relative to plays describe one form of interaction. None, alone, measures trust, relevant attention, or sales.
Normal performance is not the sponsored baseline
A creator's ordinary content record is useful, but it is not a promise about an advertisement. Virev compares sponsored posts with the same creator's usual performance and finds different patterns across tiers.
In its Instagram results, nano sponsored reels retained 0.97 times the usual views, while mega sponsored reels retained 0.77 times. The corresponding likes-and-comments ratios were 0.91 and 0.50. These are descriptive results within the source's sample and detection rules, not causal estimates of what sponsorship does.
Sponsored posts were identified using paid-partnership labels or advertising tags in captions. That can miss undisclosed advertising and include some mislabeled content. The source reports manual checks, but classification remains imperfect. A clean-looking sponsored-versus-normal comparison does not remove that uncertainty.
Content differences matter too. An advertisement may use a different format, topic, product, or posting schedule. Larger creators may work with different brands and constraints. Public observational data cannot isolate all of those choices from account size.
The practical lesson is narrower and more useful than "small creators do not lose trust." Ask for relevant sponsored examples as well as ordinary content. If the proposed price rests on an account's biggest organic hit, the team should know whether that hit resembles the work being commissioned.
A portfolio changes the question
The study's 100-nano Instagram portfolio had a modeled median cost of $11,150 and median plays of 969,000. The one-macro comparison had a modeled median cost of $8,240 and median plays of 144,000. The small-creator portfolio generated more total plays in 91% of the paired draws.
Those amounts are not equal budgets. They are not creator invoices either. The costs come from published rate tables mapped to follower counts, not from the actual amounts paid for each historical post. A pitch that presents the comparison as a live procurement quote would be misleading.
What the simulation does illustrate is the difference between evaluating a typical single post and evaluating a group. A small account may produce modest plays most of the time, while a few posts travel much farther. Combining many accounts changes the distribution of total outcomes.
That does not make the campaign predictable in every important sense. In the typical 100-creator draw, the top three posts contributed 48% of total views. A portfolio can spread execution across more creators while still concentrating its results in a few posts.
The source's 10th-to-90th-percentile range for the nano portfolio was 576,000 to 4.5 million plays. This is a wide historical simulation range, not a confidence interval or a forecast for a selected brand. It should make a planner more interested in downside and variability, not more comfortable promising the median.
Plays are not unique reach, and unique reach is not demand
Virev explicitly says it could not measure audience overlap. Its public data contains post counts, not the information needed to deduplicate everyone who watched several creators. The summed totals may count the same person more than once.
That limitation is central to the portfolio decision. One hundred creators in a tightly connected niche may repeatedly reach a similar audience. Creators in different places or communities may add more distinct viewers. Without measurement, both are hypotheses.
Repeated exposure is not necessarily waste. A considered purchase may benefit from several useful explanations. But frequency and new reach are different benefits, and a team should know which one it is buying. Adding plays cannot answer that question.
The next step in the funnel matters just as much. A campaign can generate substantial viewing among people outside the target geography, price range, or purchase occasion. Efficient plays can coexist with weak commercial value.
Before comparing tiers, define the target outcome. If the task is broad awareness, relevant distribution may be the first concern. If it is product consideration, the quality of the explanation matters. If it is immediate revenue, conversion, margin, attribution, and incremental demand enter the decision. A view-cost table cannot substitute for those questions.
Published rates are a starting assumption, not the price of the work
The study uses 56 published Instagram price tables and 37 TikTok tables. That gives it a consistent way to attach estimated costs to many accounts. It also creates a limit: real deals depend on much more than followers.
Deliverables, exclusivity, usage rights, paid amplification, revisions, production demands, deadlines, and product categories can all change the quote. An account that fits the audience unusually well may command a different price from the table. A larger package may include work that a single-post estimate leaves out.
The source attempts to remove likely boosted content from cost calculations using a performance heuristic. That is not the same as inspecting actual media spend. Some organic outliers could resemble the rule, and some paid distribution could remain undetected. The resulting numbers are model-dependent estimates.
Teams should therefore use the research to improve questions, not to dictate what a creator must charge. A fair negotiation begins with a clear scope and a verified quote. A low estimated cost per play does not establish a low total campaign cost.
Gifted products are not costless either. They have product, shipping, handling, and staff costs, and they do not guarantee a post. Where a material connection exists, disclosure obligations may also apply. The FTC's US guidance makes clear that benefits such as free products can matter; teams should check the rules for the relevant market.
The operating cost belongs in the comparison
A single creator and 100 creators require different amounts of work. More contracts, product shipments, approvals, questions, payments, and follow-ups may sit behind the small-creator option. That work is part of the buying decision, even when it is paid through salaries rather than the campaign invoice.
The Virev article includes an illustrative overhead calculation. It is useful as a reminder, but a company should replace it with its own process costs. One hour per creator is an assumption, not a measured universal staffing requirement.
A team launching a product with complicated claims may need more review than a team promoting a simple local experience. A time-sensitive release may value one reliable delivery date more than a broad portfolio. A brand with clear contracts and an established fulfillment process may carry a larger group more easily.
These differences do not invalidate the small-creator approach. They explain why the best answer is conditional. A low content fee can be expensive to coordinate. A high fee can include useful production and rights. Compare the complete scope before deciding that one tier is cheaper.
Recent BCG analysis of creator marketing also treats briefs, collaboration, and organizational choices as part of performance. It does not independently validate Virev's simulation. It supports the broader point that the result is shaped by how a company works with creators, not only which follower band it buys.
Platform and niche can change the answer
The source does not report the same single-creator economics on Instagram and TikTok. In its modeled TikTok comparison, a typical nano creator had a higher cost per thousand views than a typical mega creator. That result alone makes a universal "smaller is cheaper" rule hard to defend.
The two platforms should not be treated as interchangeable inventory. A creator's follower count, distribution pattern, content format, and relevant sponsored history need to be assessed on the platform where the post will run. Success on one account does not automatically transfer to another.
Niche comparisons also require caution. Virev's niche analysis is based on the creators whose categories it could identify, and some large-creator niche groups are small. A broad database can still contain thin evidence for a narrow purchasing decision.
For a local product, the useful pool may be much smaller than the headline account count. Language, location, category fit, availability, and conflicts can remove many otherwise attractive accounts. The operationally feasible portfolio is the one that remains after those checks, not the one implied by a global tier total.
None of this makes the research less useful. It locates the point where a benchmark stops and campaign-specific work begins. Use the benchmark to identify a question worth testing, then verify whether that question applies to the audience and creators actually available.
Older research offers support and a useful counterweight
The idea that smaller creators can deliver better commercial efficiency is not new. Revenue Generation Through Influencer Marketing, published in the Journal of Marketing, examines direct-to-consumer sales data and three field studies of paid endorsements. The authors find advantages for low-followership targeting on the financial measures they study.
That work is valuable because it goes farther down the funnel than public plays. But it belongs to specific DTC settings, definitions, and research designs. It is older evidence, not a validation that every 2026 category should buy nano creators.
Another Journal of Marketing paper, Finding Goldilocks Influencers, reports an inverted U-shaped relationship between follower count and engagement with sponsored content. Its field data covered 802 Instagram campaigns, supported by additional studies, and came from an earlier platform period.
These results should not be forced into a single ranking. They concern different outcomes and settings. Engagement volume, revenue efficiency, and modeled cost per play can favor different choices. That is a reason to define the campaign goal precisely, not to choose whichever paper supplies the most convenient headline.
A better test starts with the same decision conditions
Consider a hypothetical brand comparing a small-creator group with a larger creator for a new product. No result is assumed here. The useful first step is to write down the common conditions: audience, geography, deliverables, time window, commercial claims, rights, and the outcome the budget is meant to produce.
Then collect real quotes and relevant sponsored examples. Estimate staff and fulfillment costs. Keep paid distribution separate from organic performance. If the options have different budgets or rights, show those differences rather than making a tidy but false side-by-side comparison.
A pilot can test execution as well as content. Did the posts arrive on time? Did the creators understand the product? Were disclosures clear? Did the audience ask relevant questions? How much staff time did the group require?
For financial goals, define measurement before the posts go live. Codes and links can help track response, but attributed purchases are not automatically incremental sales. Where a causal claim matters, use an appropriate test design and state its limits. Where that design is not feasible, report the evidence without upgrading it into certainty.
Finally, examine the distribution. Did one post carry most of the result? Were the weakest posts still useful? Did the group repeatedly reach the same audience? A strong total can hide a fragile plan, while a modest total can contain a repeatable creator relationship worth developing.
Where Flonci fits
For Flonci's readers, the research points to a decision-before-spend discipline. The question is not whether small or large creators are inherently better. It is whether the proposed mix, price, audience, and work can be explained against the intended outcome.
A good decision names its assumptions. It distinguishes public counts from verified audience information, modeled rates from quotes, and historical results from a forecast. It also leaves room for the creative reason a particular creator is right for the product.
The value of a creator cannot be reduced to one favorable ratio. Neither can the risk. A team should be able to explain why the choice makes sense before the posts are commissioned, and what evidence would make it change the plan.
Conclusion: buy a plan, not a follower tier
Virev's October study challenges the habit of judging small creators one at a time while judging large creators by their apparent scale. Its simulations show why portfolios deserve attention. They also show why the cost model, concentration of outcomes, and unknown overlap belong in the same conversation.
The study does not prove a universal small-creator advantage in sales. It does not supply your next quote, your target audience, or your operating capacity. Its strongest contribution is a better way to ask the budget question.
Compare normal and sponsored performance. Use the right denominator. Price the actual scope. Include the work behind the posts. Measure the outcome that matters. Then choose the creator mix that fits the campaign rather than the tier that wins a headline.
Know before spend. And know which parts of the decision are still assumptions.
Research notes and sources
Virev, Are micro and nano influencers worth it?, October 5, 2026. Primary source. Commercial vendor observational database study with published-rate cost estimates and 10,000 random campaign draws. Public plays are not unique reach. Prices are not actual creator payments. Audience overlap and sales were not measured.
Beichert, Bayerl, Goldenberg and Lanz, Revenue Generation Through Influencer Marketing, Journal of Marketing, 2024 issue. Older supporting evidence from DTC revenue data and paid-endorsement field studies. Not pooled with the Virev data.
Finding Goldilocks Influencers: How Follower Count Drives Social Media Engagement, Journal of Marketing. Older research on sponsored-content engagement, including 802 Instagram campaigns. Different outcome and platform period.
BCG, The Power of Authenticity, October 1, 2026. Supporting context on creator collaboration and operating choices, not an independent replication of the primary study.
FTC, Endorsement Guides: What People Are Asking. US guidance checked for the article. Rules depend on the connection, format, audience, and jurisdiction; appropriate local review remains necessary.
No comments:
Post a Comment