99.8% of brands do influencer marketing wrong. We spent €4M+ finding out why.

I'm Santiago, founder of HappyFlops. I discovered this strategy together with my partners after spending over €4M on influencers. We made the expensive mistakes, found the patterns underneath them, and today we hit an 89% hit rate when we choose who to work with. I'm writing this for one reason: so you don't burn your money on influencers that will never convert.

You were told influencer marketing is simple: find someone with a big following, pay them, get sales. So you did. Maybe it worked once. Mostly it didn't, and the worst part is you never found out why. The money just disappeared into a post that looked fine and did nothing.

For most brands, influencer marketing is a casino. You pay a lot, hit publish, and pray it works. Sometimes the machine pays out. Usually it doesn't, and you walk away a little poorer and none the wiser.

Here's why. Most people think about engagement rate, followers and reach. That is the whole problem. Those numbers are not what move real sales. The account with the biggest audience is almost never the one that sells the most product, and we have the data on 200+ of them to prove it.

What actually sells: trust

The one thing every creator who prints money for a brand shares isn't size. It's trust: the specific belief their audience holds that when this person recommends something, it's real.

But how do you measure trust? It sounds impossible. For years we thought so too. Then we found a way to put a real number on it, using actual data instead of gut feeling: four different methods that each measure a different signal of trust. On their own, each is useful. Combined, they give us an 89% hit rate on picking talent that converts, before we spend a single euro.

We didn't arrive at that overnight. We tested 200+ celebrities and influencers across 8 markets with €4M+ of our own money, tracked exactly which ones converted, and watched the pattern emerge. Out of all 200+, roughly 20 became the winners that carry the results. Those four methods are simply how we now spot the 20 in advance.

And it works at both ends of the scale. The same signals tell us whether a 1M+ follower celebrity will actually move product, or whether a small, almost unknown creator will quietly outsell them. Followers are just one input. More often than people expect, the smaller name wins.

Why this works even harder now

The internet is drowning in AI content. Everything looks polished, everything sounds the same, and buyers have stopped believing most of it.

That's exactly why real, trusted human endorsement is worth more today than it was five years ago. It's one of the last things that still cuts through. The brands that understand this are quietly compounding while everyone else fights over ad costs.

We own the brands. We're not an agency.

We're not teaching from the sidelines. We scaled HappyFlops from €0 to €80M+ in three years using exactly this, and the same system sits behind €88M+ of brands we own. We've since signed 7 brands in 7 different niches and done the same for them. It's the system, not the niche. And it doesn't take a big budget: we've booked profiles under €3k that generated €100k+ in sales. Cheap isn't the problem. Guessing is.