Why Algorithms
Don't Build Brands.
People Do.

There is a recurring conversation I have with brand teams. It usually begins with some version of: "We need to crack the algorithm." And it usually ends the same way — with a strategy that's optimised for reach and utterly invisible to the people it reaches.

The algorithm is not the audience. It never was. It's a distribution system. And distribution systems do not make people feel things.

Here's what actually happens when a brand chases the algorithm. They study what formats are trending. They replicate them without understanding why they worked. They post at the right frequency. They hit every platform. They end up with a content calendar that looks like strategy and functions like wallpaper.

The brands that win aren't the ones that figured out the algorithm. They're the ones the algorithm had no choice but to serve — because people kept engaging with them, sharing them, and searching for them. The algorithm is a response mechanism. It responds to human behaviour. Build for the human first.

The three things that actually build brands

I've spent the last several years watching brands at both ends of the spectrum — the ones that grow into institutions and the ones that grow into noise. The difference is not budget. It's not even category. It's almost always one of three things.

• They have a genuine point of view. Not a tagline. An actual perspective on the world that shapes every decision. Biryani By Kilo didn't compete on speed or price. They made biryani a ritual. That's a point of view. Everything else followed from it.
• They build cultural fluency, not just social presence. Cultural fluency means you understand what your audience cares about at a cultural level and how your brand can genuinely participate — not just reference. A billboard made of actual receipts that goes viral because it captures exactly how people feel about spending money — that's cultural fluency.
• They earn attention before they ask for it. The best-performing content we've ever made was the content that gave people something before it asked them for anything. A laugh. A feeling. A perspective they hadn't considered. The ask — the conversion, the sale — came later, when it was earned.

What the numbers actually say

If you want to make this argument in a room full of performance marketers, you'll need data. Here's what we've seen across our portfolio.

These numbers aren't interesting because they're big. They're interesting because of why they're big. The ROAS didn't come from better bidding. It came from better creative — creative that resonated with real people, reduced the resistance to purchase, and turned first-time buyers into repeat customers.

What to actually do about it

None of this means ignore the algorithm. It means use it correctly. The algorithm is extraordinarily good at finding more people who look like the people already engaging with you. So the question is: what are they engaging with, and why?

Build the thing that earns real engagement first. Understand what your audience cares about at a cultural level — not just a demographic level. Give them something worth sharing before you ask them for anything. Then let the algorithm do what it's actually good at: scaling something that already works.

The brands that last aren't algorithm-optimised. They're human-resonant. The algorithm will catch up. It always does.