The most repeated piece of marketing advice this year is that brands should behave less like advertisers and more like creators. It sounds like a note about tone. It's really a response to an economic shift: production got cheap for everybody, so production stopped being an advantage.
The numbers underneath it
About 64% of Gen Z say humor makes them more likely to share branded content. Around 61% trust creator recommendations over traditional brand advertising. Roughly 68% say authentic, unpolished content beats highly produced ads.
Put those together and you have an audience that reads polish as distance. A flawless spot announces a budget and a committee. A funny, slightly rough post announces a person.
Duolingo is the example everyone reaches for, and it deserves the mention. The brand went Gen Z-native by acting like a creator instead of a corporation — irreverent, self-aware, at home on the platform it posts to. What it built is a character people recognize, not an advertiser interrupting one.
Why polish stopped working
Generative tools flattened the cost of making things that look professional. When anyone can turn out clean, competent, on-brand assets at volume, competence stops setting anyone apart. What's scarce now is a point of view.
That recasts the humor finding. Humor isn't a trick for grabbing attention — it's evidence that a specific person with specific taste made a call, which is exactly the thing consensus can't produce. Same logic explains why visibly human work is gaining signaling value as AI-made marketing saturates every channel.
Smaller creators, more trust
The influencer market has reorganized around the same insight. Skepticism toward mega-influencers and celebrity endorsements keeps growing, while micro- and nano-creators with specific, trusting audiences outperform on engagement and conversion.
The mechanism isn't mysterious. A creator with 8,000 followers in one niche has an audience that knows their taste and notices immediately when a recommendation is out of character. That accountability is what makes the recommendation worth anything — and it's precisely what evaporates at celebrity scale.
Alongside that, fandoms and closed communities have become distribution channels in their own right: group chats, private servers, subreddits, newsletters, places where recommendations move between people who trust each other instead of an algorithm.
What this asks of a marketing team
More than most teams expect. Acting like a creator means posting consistently in a recognizable voice, reacting to what happened this week, and accepting that some posts will flop because a voice with range occasionally misses.
That runs straight into how brand approval usually works. Committee review sands off the exact specificity that makes the whole approach function, and campaign-cycle planning can't respond to a cultural moment three days late. The teams getting results have generally pushed publishing authority down to the people who actually understand the platform, with guardrails instead of sign-offs.
The practical version
Use AI for variation and adaptation — resizing, versioning, localizing — wherever volume genuinely matters, and keep the core idea and the voice human. Move influencer budget away from a couple of big names toward more small ones with real authority in a niche. And write like someone specific, because the alternative now costs nothing for anyone to make, and it's worth about that much.
Image: Anna Shvets, via Pexels
