Distribution Ate Everything
OpenAI paid hundreds of millions of dollars for TBPN. A podcast. Two guys talking into microphones.
The most advanced AI lab on the planet paid more for a podcast than 99% of venture-backed startups will ever exit for.
a16z is building media arms. Paramount grabbed The Free Press for $150M. Stripe swallowed Indie Hackers. HubSpot bought The Hustle. Semafor raised $30M at a $330M valuation. Fox took Red Seat Ventures.
The smartest money on earth is making the same bet. And it's not on products.
It's on eyeballs.
A 14-year-old with Claude can ship an app in a weekend. Multiply that by a few million kids with laptops and wifi. They can build infinite apps, infinite content, infinite AI agents flooding every platform nonstop.
You can't print new eyeballs. There are 8 billion humans and 24 hours in a day. That's it. That's the global supply of attention. Production costs are collapsing to zero but CPMs keep climbing.
This is the distribution supercycle.
The old playbook is cooked
Meta CPMs crossed $40 in most B2C niches. Customer acquisition cost is up 222%. The average app converts at 1.2% on paid traffic. For every 1,000 people you pay to click, 988 ghost you.
Zuckerberg's new pitch is "give us your URL and we'll do the rest." When Meta's AI decides who sees your ads, creative quality and your willingness to stomach higher CPMs are all you have left. Their own data showed cost per conversion dropped 22.6% when they removed manual targeting. Good for their algorithm. Terrible for founders who relied on targeting precision to compete with bigger budgets.
When ads run through the same AI, the company that can pay the most per impression wins. Startups are not great at budget wars.
Billboards, TV, radio, print. Priced in 1998 dollars, performing to 1998 attention spans, still charging 2026 rates. The people still buying them confuse visibility with distribution.
Traditional influencers. $4,000 for a video averaging 3,000 views. Flat rate for a coin flip. Every one of these channels charges you the same whether the campaign prints money or dies in silence.
What actually works
Organic distribution. There are only four ways to play:
- Algorithmic text (X, Threads, LinkedIn, IG carousels)
- Algorithmic short-form video (Reels, TikTok, Shorts, Facebook Reels, LinkedIn video)
- Algorithmic long-form video (YouTube)
- Long-form video intentionally made to be clipped into short-form
That's it. If you don't have someone in your org who has genuinely figured out one of these four routes, you will get outrun by someone who has.
Kevin O'Leary used to pay customer acquisition managers $48,000. He now pays them $250,000. Because you can measure their output by revenue generated per week. Some of them go contractor and clear half a million a year. They know how to take content, cut it into a 59-second ad, and acquire 200 customers. Those people, usually in their early twenties, are the most valuable hires in business right now.
Clips are the new Facebook ads
Content distribution runs on clips now. Podcasts are clip factories. The average successful show produces 30 to 50 clips per episode and those clips reach 20x to 50x the audience of the full episode. Most fans consume the clips, not the show. The show is raw material. Clips are the distribution.
Both presidential campaigns in 2024 hired dedicated social clipping teams. The Harris campaign streamlined content approvals so clips of debate moments and rally reactions hit TikTok, Shorts, and Reels within minutes. 60-second clips reached more voters than any TV ad buy in history. Short-form distribution became an electoral strategy for the presidency of the United States.
The window
This is still early. UGC distribution and clip engines are where Facebook ads were in 2008. The founders who figure this out now will own their distribution for the next decade. The ones who keep renting attention from ad platforms will spend more every quarter for less.
The product is easy to copy. Distribution is the thing that compounds.