The 2026 Edelman Trust Barometer, released in January from interviews with over 17,000 people across 15 countries, put a number on something marketers had only felt anecdotally: seven in ten people are now insular. They are hesitant or unwilling to trust anyone who holds different values, gets information from different sources, or comes from a different cultural background. Three months later, TikTok's ranking system quietly started doing the same math. Accounts that publish across more than three unrelated topic clusters now lose roughly 45% of their reach compared to accounts that stay inside one. The platform never read the Trust Barometer. It reached the same conclusion from engagement data alone.
Why did a ranking algorithm and a global trust survey converge on the same rule?
They are measuring the same behavior with two different instruments. TikTok's recommendation engine clusters viewers into small interest groups and tests whether a video holds them there. A creator who posts cooking one week and finance the next confuses the clustering, so the system can't confidently place the account with an audience and shows it to almost no one. Edelman's survey found the human version of that same confusion-avoidance: when someone doesn't share your values or your information diet, the brain treats them as noise rather than signal, and disengages before persuasion gets a chance to work. Algorithm and audience are both routing around anything that doesn't cleanly belong to one group.
What happens to a media plan built for reach when both the code and the customer punish it?
It gets expensive twice. First, paid reach into an insular audience is simply worth less. Edelman found unpaid voices (friends, family, "someone like me") are five times more powerful than paid brand messaging at building trust with insular consumers, and only 46% of that group say any brand voice moves them at all if it isn't already inside their circle. Second, the delivery mechanism itself now costs more to run broad. A campaign spread across a creator's mixed content, or a brand account posting to multiple audience segments, is fighting a structural 45% reach tax before the creative is even judged. You're paying list rates for a lever the platform has quietly weakened.
So does going broad still work at all?
Rarely, and only where the category is genuinely universal. For most categories, the same Edelman report found 84% of consumers say they need to share values with a brand before buying from it. That figure has held steady since 2024, which tells you this isn't a passing mood. It's a baseline platforms have only just caught up to mechanically. Add in earlier 2026 data showing 53% of consumers now distrust paid influencer endorsements outright, while 92% say they trust nano- and micro-creators more than mega-influencers, precisely because those accounts read as belonging to one identifiable group. The instinct to buy fewer, bigger, broader placements is fighting the exact mechanism, human and algorithmic, that decides whether anyone sees or believes the message at all.
- Audit your reach premium. Any placement priced for breadth is now competing against a structural delivery discount most buyers haven't priced in.
- Fund fewer, narrower, longer. A nano-creator embedded in one cluster for 60-90 days builds the algorithmic trust a broad campaign can't buy in a flight.
- Route budget through "someone like me," not celebrity. The insular distrust celebrity endorsement at nearly double the rate they distrust a peer.
- Stop treating category-fit as optional. Values alignment isn't brand positioning anymore. It's the gate that decides if the algorithm shows your message to anyone.
The platform stopped rewarding reach and started rewarding belonging, and it did it roughly the same season the world's largest trust survey said people had already made that switch themselves.
What this means for your next media plan
Run the arithmetic before the next renewal. One broad placement bought at premium CPMs, delivered at a structural reach discount, and received by a skeptical, insular audience is now competing against ten narrower placements. Each one is embedded in a single niche cluster, each runs the 30-to-90 day window TikTok's system needs to trust an account's identity, and each is carried by a voice the audience already coded as "someone like me." The second bundle likely costs less in aggregate and reaches people who were actually going to believe it. If your 2026 budget still allocates as though reach were the scarce resource, it's optimizing for the world Edelman just measured us out of.
Want this kind of thinking on your brand?
We build brand strategy, AI content and performance for the AI era.
START A PROJECT →