Trust in public-facing authority is collapsing. Creator marketing budgets are exploding. Nobody in the industry seems to notice those two facts don't fit together.
This topic was brought to our attention through The Modern Strategist newsletter, where readers send in the arguments their own teams can't settle. This one arrived exactly that way, one side certain the money should follow creators, the other certain that metric stopped meaning anything, neither able to produce a number that actually settled it. So we tested it.
Influencer Marketing vs. Word of Mouth: What the Data Says About Purchase Behavior in 2026
The Budget Side Is Not Subtle
Creator spend is not plateauing. It is compounding. According to Aspire's State of Influencer Marketing 2026, US influencer marketing spend is projected to grow 15.7% in 2026, on its way to $13.7B by 2027. Brands are now putting an average of 23% of total marketing budget into creator partnerships. That is not a test-and-learn allocation. That is a primary channel.
The confidence behind that spend is close to unanimous. Influencer Marketing Hub's 2026 Benchmark Report puts it plainly: 87.49% of brands expect to increase influencer budgets this year. Just 5.55% expect to cut them. When 87% of an industry agrees on anything, that agreement is usually treated as the evidence. Nobody asks what it is actually evidence of.
Budget confidence and consumer trust are not the same measurement. One tells you what marketers believe works. The other tells you what consumers say moves them. An entire industry can be confident in a channel that a growing share of its target audience is quietly discounting, and both numbers can be accurate at the same time. That is not a paradox. It is two different surveys answering two different questions, and treating them as interchangeable is exactly how a budget gets misallocated for years before anyone notices.
The Trust Side Tells a Different Story
At the same time budgets are scaling into public-facing creators, consumers are pulling trust in the opposite direction, toward the people they already know.
The 2026 Edelman Trust Barometer calls it “From We to Me.” Trust in neighbors, family, and friends is up, net +11. Trust in national government leaders is down, net -16. Trust in major news organizations is down, net -11. The pattern is not subtle either: people are retreating from public, institutional, broadcast-style authority and consolidating trust into small, personal circles.
Put those two data points next to each other and the budget logic gets shaky. Brands are spending more than ever on a form of public-facing influence, at exactly the moment consumers say they're trusting public-facing influence less. Something has to give. Either the spend is wrong, or the trust data does not mean what it looks like it means.
The Wrinkle That Keeps This Honest
Here is where it gets genuinely undecided, not a strawman we get to knock down in three weeks.
The same Edelman research complicates its own headline. Among people who say they trust a food or lifestyle influencer, 48% of respondents, 62% say they would trust, or consider trusting, a company they currently distrust, if that influencer vouched for it. Among the smaller group who trust a financial influencer (44% of respondents), the number is 57%.
That is not a dying channel. That is a conditional one. Influencer trust is not gone, it is narrower and more specific than a follower count or an engagement rate can show. It travels through people who have already earned trust in a category, and it moves consumers who were already primed to listen. Raw reach was never measuring that. It was measuring something adjacent to it and calling it the same thing.
The Diagnosis That Determines the Strategy: Trust vs. Behavior
Here is the version of this argument the industry has not tested: what a consumer says moves their trust, and what actually shows up as the last thing they interacted with before a real, recent purchase, are two different datasets. Almost everything cited above, including our own sources, measures the first. Stated trust. Budget confidence. Sentiment.
Almost nothing measures the second. The gap between what moves someone's stated trust and what actually sits in the final seconds before they buy is where this entire argument is either won or lost, and until now it has been unmeasured. That gap is what most budget decisions are quietly assuming away.
Standard Insights surveyed 507 US adults, nationally representative, to close it. We built the survey around one central diagnostic: forced against each other, does a friend's recommendation actually beat an influencer's, or does the industry only assume it does not because nobody has run the head-to-head?
What We Tested - and What 507 Consumers Actually Said
Ask someone directly whether a friend or an influencer moves them more, and it is a coin flip. Force them to actually choose between the two, and it is not. Not even close.
25.0% say influencers carry more purchasing influence. 24.9% say friends and family. 33.1% call it about equal. That is a statistical tie. Then we forced the choice: pick a friend or an influencer, when both are recommending competing products. 31.4% picked the friend. 2.2% picked the influencer. Among everyone willing to take a side, friends win 93.5% of the vote.
Stated belief and forced behavior are not the same measurement. Only one of them predicts what people actually do, and it is not the one the industry keeps asking about.
The Source Positioning Matrix: Six Channels, One Pattern
We mapped six recommendation sources on two dimensions: how much consumers say they trust each one, and how much each one actually drove a real, unplanned purchase in the last three months.
The pattern is not subtle. Every single influencer tier delivers less actual purchase behavior than its stated trust would predict. Brand advertising does the opposite, it delivers nearly six times the purchase behavior its trust score would suggest, the most under-trusted, over-performing source in the entire dataset.
That gap is the edition's real finding. Consumers say they trust brand advertising least of anything we tested. Then it outperforms every other source, including the people they say they trust most.
The Channel Getting No Credit
We asked what specifically led to a recent purchase that consumers describe as recommendation-driven. Friends and family: 37.5%. Influencers, combined across all three tiers: 44.8%. Brand advertising: 2.8%.
Brand advertising drives more unplanned purchases than any single source in this dataset, 55.4%, while being almost never credited as a recommendation. That is not a contradiction. It is the finding. Brand ads do not get remembered as advice. They work anyway, quietly, beneath what consumers believe is actually moving them. The channel doing the most is the channel getting the least credit, including from the people it is working on.
The Buyer Gap: Active vs. Passive Social Media Users
Active social media users, people on platforms multiple times a day, behave nothing like passive ones. 94.2% of active users have bought something via an influencer. Among passive users, that is 46.0%, a 48.2-point gap, the widest in the dataset.
Recent purchases attributed to an influencer: 62.7% of active users versus 20.5% of passive users, a 42.2-point gap. Belief that influencers outweigh friends and family: 39.7% active versus 5.1% passive, 34.6 points apart.
That is a platform-immersion divide, not an age divide. It splits the population far more cleanly than generation does, which the segment data below confirms directly.
Trust by Age, Income, and Gender: The Signals Marketers Are Missing
The Generational Story Doesn't Hold
The industry has been treating niche-creator trust as a young-consumer phenomenon. The data says no. Trust in micro/nano influencers sits at 14.8% among 18–34-year-olds and 14.5% among consumers 55 and older, functionally identical. Among Gen Z specifically (18–24, n=67), the single most trusted source is still friends and family, at 28.4%, the same top choice as every other cohort in the survey.
This is not a hedge on the generational narrative. It is a correction to it. “Younger consumers trust niche creators more” is not what this dataset shows, at any age cut we tested.
Regret and Income: Read the Buyer-Adjusted Number, Not the Raw One
Across the full sample of influencer buyers, 38.5% report regretting a purchase (144 of 374). Regret itself is roughly flat across income brackets, sitting between 27.0% and 29.1% of each bracket's total sample. But adoption is not flat at all, and that is what makes the raw comparison misleading.
Consumers earning $200,000 or more have the lowest influencer adoption of any income tier, 30.2% have never bought via one. Consumers earning $100,000–$199,000 show the opposite pattern: 75.4% adoption, the strongest of any bracket, combined with the lowest buyer-adjusted regret rate at 35.9%. That is the group converting most and regretting it least. Any cross-bracket comparison that does not adjust for adoption rate will overstate regret in high-adoption brackets and understate it in low-adoption ones.
One more segment worth a sentence, not a section: among people who have bought something via an influencer, male buyers report a 43.2% regret rate, compared to 33.9% among female buyers, a 9.3-point gap.
The Strategic Verdict for Marketing Leaders
Winner: brand advertising. It is structurally undervalued by marketers' budget confidence and by consumers' own sense of what moves them. It drives more actual unplanned purchases than any other source in this dataset while earning almost no conscious credit for doing so. The budget confidence data we opened with, 87.49% of brands increasing influencer spend in 2026, is chasing tiers that, without exception, convert stated trust into purchase behavior worse than the channel nobody is arguing about.
Warning signal: the celebrity/mega influencer tier. It posts the lowest purchase-influence-to-trust ratio of any source we tested, at -3.9. It earns the least recommendation credit of the three influencer tiers, 15.4%, despite carrying mid-pack trust at 14.6%. And its reach looks broad but behaves narrow, its purchase influence is concentrated almost entirely inside the active-user bubble, not spread across the general population the way its follower counts imply. It is also the tier easiest to overfund on reach alone: 2026's projected influencer budget growth, 15.7%, en route to $13.7B by 2027, rewards exactly the kind of broad-follower-count logic this data says does not convert. The data says celebrity/mega is the weakest converter of the six sources we measured.
The fight we opened with assumed the real axis was creators versus word of mouth. It was not. The real axis was stated preference versus actual behavior, and the channel neither side of that argument was even discussing, brand advertising, is the one winning on the metric that actually matters.
Want the full breakdown? The complete report includes the full six-source positioning matrix, the buyer-gap analysis between active and passive social media users, and the income and gender signals that tell you where regret actually concentrates. Read the full interactive report.
Want full workspace access, the AI chat, the live filters, the underlying dataset? Get in touch and we will add it to your workspace, free.
Methodology: 507 US adults, nationally representative. Active Social Media Users (multiple times a day) N=292, 57.6%. Passive/Occasional Users (once a day or less) N=215, 42.4%. Survey conducted by Standard Insights, July 2026.