Influencer marketing has failed.

For a McDonald’s launch or a fast-fashion drop, an influencer campaign still moves a number on a dashboard. For a small restaurant with a real point of view, the same playbook has spent a decade quietly delivering the wrong customer.

The reach metrics looked fine. The room felt off. That mismatch is not an opinion; it’s the lived experience of running a culture-forward venue and watching paid creator content land like an ill-fitting suit.

The thesis of this piece is simple. Influencer marketing was promised as authentic word-of-mouth at scale. It mutated into rate-card content that optimises for sponsorship volume. And it is now being quietly replaced by something better: a distributed network of small voices who post because they care, not because they were paid to. The shift matters most for small, opinionated venues with a tribe, and barely matters at all for commodity brands. We’ll explain why.

The promise

In 2010, the idea was hopeful. Legacy media gatekeepers — newspapers, glossy magazines, the friendly TV restaurant reviewer — had spent decades deciding which restaurants, bars, and brands existed in the public imagination. Most of them never got to. A handful of cities, a handful of voices, a handful of opinions on rotation.

The early social-media generation looked at this gatekeeping machine and saw something obviously wrong. There were thousands of cooks, writers, drinkers, photographers, and neighbourhood obsessives across the world with sharper eyes than the legacy critics. They just had nowhere to publish. The platforms — first blogs, then Twitter, then Instagram, then TikTok — were going to fix that.

The promise of early influencer marketing was that voices who didn’t owe a magazine or a TV network anything could reach as many people as a magazine could, faster and cheaper, with opinions formed inside the scene rather than from a press junket. The relationship between a voice and its audience would feel closer to friendship than broadcast.

For culture-forward restaurants and bars, this was supposed to be the deal of the decade. A venue with a thesis — a Hokkien-inspired tasting menu, a natural wine list, a hi-fi listening bar — had finally found a marketing channel that could match its own voice. The independent creator was the modern-day word-of-mouth. The recommendation network was finally democratised.

For about three years, the promise mostly held. Then it broke.

The betrayal

What changed was money, and then what money does to incentives.

By 2016, the term “influencer” had stopped meaning “person with a point of view” and started meaning “person who could be paid to post.” A whole industry (agencies, talent managers, rate-card consultants, sponsorship spreadsheets) assembled itself around monetising the audience that had been built on authenticity. What the industry sold to brands was reach. What it took from creators was, eventually, their credibility.

The mathematics of this were predictable. If you have an audience and a rate card, the rational thing is to fill the schedule. If you fill the schedule with sponsored posts, your audience starts seeing more sponsored posts than genuine recommendations. Your audience adjusts and trusts you less. You compensate by being louder, more positive, more performatively excited. Trust erodes faster. You compensate again.

We’ve all sat through the result. A “creator” posts a glowing video about a restaurant on Monday, a glowing video about a competing restaurant on Wednesday, a glowing video about an unrelated tech product on Friday. The implicit message, I like everything, is of course the opposite of a useful recommendation.

A recommendation only works as a recommendation when there is something the recommender wouldn’t say yes to.

There’s a second, subtler failure that affects even the most carefully selected influencer placements: the credibility of any single recommendation is now a function of the recommender’s overall feed, not the placement itself. If your favourite food creator’s last six posts are sponsored, even thoughtful and well-disclosed ones, the seventh unsponsored post about a new restaurant doesn’t read as a recommendation. It reads as an unsponsored post that will probably be followed by a sponsored one.

The problem is not any individual creator. It is the saturation of the medium. The operators we talk to feel sponsored-content fatigue accelerating. The brands that need credibility most, the ones whose value proposition isa specific sensibility, are paying for placements in a channel that has spent ten years burning down the trust they’re trying to borrow.

Why this hits culture-driven venues hardest

A restaurant or bar with a thesis is not selling food or drink. It is selling a sensibility: a specific, calibrated set of choices about taste, music, lighting, service, and what a good night out looks like to the people who built it. The audience it deserves is the slice of the city who shares that sensibility. The audience it does not deserve is everyone else.

Influencer marketing, as currently practised, is structurally incapable of distinguishing between these audiences. A 60-second TikTok with the right hashtags and a paid push reaches whoever the algorithm decides to reach. Reach is the metric, and reach is indifferent to taste.

What it delivers, in practice, is the wrong customer. People who came because TikTok said it was trending sit through service confused, leave half-good reviews, and tell their friends it was “fine.” The attention comes in a flood for a week or two and then falls off a cliff when the algorithm moves on. There’s no compounding effect because there was no relationship to begin with.

And then there’s the slower damage. A venue with a real tribe, the kind that comes back and brings friends, is built one decision at a time. Filling the room with people who came for the wrong reasons dilutes that tribe. The regulars feel it before the operators do.

We’ve watched this happen, repeatedly. A bakery in Tiong Bahru does the influencer rounds, gets a queue around the block, and within three months the queue has moved to the next viral spot, while the original audience, the one that actually believed in what the bakery was doing, has quietly drifted away because the experience changed to handle volume that was never the point.

The venue knows its tribe. The agency knows its price list. The two are not the same thing.

The new renaissance

Here is the good news.

A second generation of voices is emerging, and it looks nothing like the first. Not because the platforms changed, but because the people did. The new voices are filtered by a specific set of conditions that the original creator economy never had: the sponsorship market is saturated (audiences now read bidding language at a glance); the algorithm rewards specificity over generality (a niche feed compounds where a broad one stalls); and an entire generation watched the first wave cash in and decided they wanted no part of that ending. For a closer look at what these voices actually look like on a feed, see How to find credible food creators.

What you get from those filters is people who post about places because they care, with full-time jobs that pay their rent, on small accounts with audiences they actually know. The economics don’t pull them toward the rate card, because the rate card isn’t the prize anymore. They are:

They’re smaller, sometimes only a few hundred to a few thousand followers. Their feeds aren’t their job; the feeds run alongside something else paying the rent. They post about one neighbourhood, one obsession, one category of drink, and the audience is small because it’s filtered to the people who care about that exact thing. They post less often and the signal-to-noise stays high. Most of them have never taken sponsorship money and don’t want to.

These are the people we call tastemakers. The label matters: they are not influencers, and treating them like influencers (paying them, briefing them, scripting them) would immediately collapse the thing that makes them credible. The first wave of the creator economy taught us this the hard way. (If you’re a creator wondering whether you fit the description, the criteria we use are written out in What we look for in a tastemaker.)

What this looks like in practice

The shift from influencer marketing to tastemaker networks isn’t theoretical. Three operational moves carry most of the difference:

Instead of paying a creator, you invite them. A hosted experience, a tasting, a soft launch, an early look at a new menu — not payment, an exchange of access for honest public reflection. Credible voices say yes happily; the professional rate-card people decline because there’s no cash and no clout, which is the filter working as intended.

And the question stops being “how many people will see this post.” It becomes how many of the right people will, from a voice they already trust. The numbers come down. The work gets infinitely more useful. A single sponsored post is an event; a network of small voices talking about your venue over six months is a culture, and only the second compounds.

You are not buying reach. You are earning conversation — a conversation that happens because credible voices were given a reason to talk, and trusted to talk in their own way.

To be specific about what we’re claiming: tastemaker networks are not a replacement for paid marketing across all categories. If you’re launching a chain of fast-casual concepts, an influencer campaign is probably still the right tool. Commodity brands can run commodity marketing and survive the dilution.

If you’re launching a single, opinionated, neighbourhood-defining venue — a thoughtful restaurant, a niche bar, a hi-fi listening room — you don’t need reach. You need the right tribe, their attention, their trust, and their friends.

That’s what tastemaker networks build, slowly, post by post, over time.

In rooms we sit in across Singapore, London, Hong Kong, and New York, the same conversation is happening among operators of thoughtful venues:

“The influencer thing isn’t working anymore. We did the run, we got the numbers, we got the wrong people. We’re not doing it again.”

The shape of the new approach is clear: small voices, small audiences, repeated over time, curated on both sides, never paid for what they say.

The promise wasn’t wrong. It was just too early, and it arrived in the wrong economic frame.