Stop Renting Attention. You Can't Afford It Anymore.

I'm standing in a slow queue at a kopitiam in PJ. The kind where the uncle behind the counter moves at his own pace and you just have to respect it.

Everyone around me is doing the same thing: head down, thumb flicking upward. I watch a woman in front of me scroll past a snack review, a celebrity skincare post, and a discount code with a countdown timer — all inside eight seconds. I can tell she registered none of it. Not because the ads were bad. Because they were identical to everything else.

That's the wall performance marketing just hit.

For a decade, the game was efficiency: buy the right click at the lowest price. But when every brand uses the same automated tools to chase the same shrinking pool of attention, the advantage disappears. You're all bidding on the same room. The moment you stop feeding the meter, the room goes silent.

Most brands have been tenants their whole lives and called it a strategy.

From tenant to landlord

Before I ran strategy, I ran productions. And before that, I studied law. The LLB didn't stick as a career, but the framework did: think about value in terms of rights and ownership, not reach and impressions.

A campaign is a sprint you pay for and then it's gone. A narrative asset sits on your balance sheet. It has exploitation lanes: theatrical, non-theatrical (airlines, events, licensing), and the streaming triad of SVOD, AVOD, and FAST. If you create something good and license it all away for a moment of reach, you've donated your future to a platform that doesn't know your name.

This is what Brand-as-Studio actually means. Not content marketing. Not a content calendar. The shift from renting culture to owning it.

Owned channels distribute — newsletter, podcast, website. Owned audiences return — subscribers, CRM. Owned IP compounds, because it's a reusable story asset that can travel across channels and still mean something two years from now.

That's the only defensible position left.

What the algorithm is actually doing

The modern algorithm isn't a follower graph anymore. It's a topic graph. It acts like a matchmaker.

When you post, it builds a fingerprint — computer vision, audio analysis, behavioural signals. Then it tests that fingerprint against roughly 200 strangers. If your content fits, it goes wider. If it doesn't, it goes to view jail and stays there.

The brands that fail this test are the ones acting like variety shows. Technical post on Monday, lifestyle content on Wednesday, product promo on Friday. The algorithm gets mixed signals. The fit score drops. Distribution collapses.

Treat your brand like a TV network. Every post is a pilot. The ones that perform become series. The series become IP. That's how you move from renting attention to owning it.

And it's not just search on Google anymore. Instagram is doing over 6.5 billion searches a day. If your brand doesn't have a consistent, narrow format that AI engines can fingerprint and verify, you're invisible to a massive part of how people actually discover things now.

The mechanics of keeping attention

There's a vertical storytelling framework I first came across in Adi Tiwary's exploration of micro-dramas and the Brand-as-Studio model — the 3.7.21 rule.

• Zero to three seconds: break the pattern visually, or you've already lost.

• Three to seven seconds: state the theme plainly — who's stuck, what's at stake.

• By 21 seconds: create a commitment spike. Make the viewer pick a side.

Beyond that, the structure I keep coming back to for longer-form social is a 90-second cliffhanger grid:

• 0–5s: Visual hook. Pattern break before the scroll continues.

• 5–30s: Rapid exposition. Stakes established.

• 30–75s: Escalation. Reversals, tension.

• 75–85s: Internal twist. Something that deepens the emotional read.

• 85–90s: Hard cut, mid-action. "Next episode" behavior triggered.

This isn't dopamine farming. It's narrative durability — the ability of a story to retain attention long after the media spend is gone.

Test before you build the factory

The smartest thing I've seen brands do in 2026: operate like social scientists. Don't bet your main brand equity on every idea. Test pilots on separate, unbranded accounts first. The "Ramen on the Street" model. A mockumentary format. A character-driven series that funnels back to the main brand once you know it works.

An 8-week cycle is enough to know:

• Weeks 1–2: Define the narrative spine. Who's the protagonist? What are the real stakes?

• Weeks 3–4: Map the cliffhanger grid. Negotiate your exploitation lanes early — merch, distribution, licensing.

• Weeks 5–6: Design re-entry hooks. Assume people will drop off and come back; make it easy to catch up.

• Weeks 7–8: Measure retention, not vanity. Hook rate at 3 seconds. Commit rate at 21 seconds. Find the outlier that's getting 10x the engagement and study it before you do anything else.

Only after that do you build the content factory. Build for what's already working.

The Geng Kubur lesson

There's a case study most SEA marketers haven't seen yet — Ford's John Bronco campaign in the US. I studied it closely. The short version: Ford didn't launch a product, they built a character with a strong narrative arc before the vehicle even shipped.

Red Bull does the same thing differently — they haven't run a traditional ad in years. They own a media company that happens to sell energy drinks.

I have my own version of this lesson:

If I could go back to how we released Geng Kubur (Dead Boys Club) in 2024, I'd apply the distributor-plus-sales-agent mindset from day one. Don't launch a movie — build a 12-month release pathway. Treat the film as a commercial asset with specific rights. We actually did negotiate multi-million ringgit in marketing exposure through our Primeworks partnership, and we did close a Netflix SEA distribution deal. But I'd have structured the rights architecture earlier. I think the box office would have looked different.

The lesson isn't specific to film. Any brand IP works the same way. Build it like you intend to sell it. Because the day you decide to sell it, the architecture either exists or it doesn't.

On AI, and what it can't do

A quick word on this, because it keeps coming up.

Brandwatch is tracking a 200% spike in mentions of "slop" — AI-generated content that feels hollow. Consumers have developed a nose for it faster than most brands expected. One researcher described it as content that's been sanitised to the point that audiences can smell it.

I've said this before and I'll keep saying it: AI handles the mechanical. Humans bring the soul.

I think about creative work the way I think about martial arts. In a 12-round fight, you can't fake the conditioning. You can't fake the thousands of hours of repetition. The audience knows. They've always known. What's changed is that in an era of blanding — safe aesthetics, zero risk, endless beige — earnestness is actually a competitive edge. Showing your work is the only way to prove you're in the ring.

AI gets you to a decent first draft faster than before. That's real, and I use it. But the specific moment, the cultural reference someone else wouldn't reach for, the thing that makes a piece actually yours — that still has to come from a human who was actually in the room.

You're not trying to become a streamer.

You're trying to stop renting attention forever. That's the whole thing.

On a final note:

I'm building a playbook for what I'm calling Permissionless IP — a framework for SEA creators and brand builders who are done waiting for platform permission, distributor approval, or any type of validation to own their story.

The first cohort gets two things nobody else will: early access to the playbook before it's published, and personal sessions with me on how to apply it to their specific IP.

I'm keeping the first cohort small, under 15 people, so I can give proper attention to each one. If this is for you, get on the list.

[Join the waitlist →] Permissionless IP — First Cohort Waitlist