Tip
Brian Balfour talks about platform cycles: “I lived the full cycle in a very short period. I lived the glory days and just the absolute horror days” with Facebook’s platform. He’s watched it repeat with Google SEO, Apple App Store, LinkedIn. The pattern: platforms open (gold rush), you grow like crazy, then platforms close for monetization and you’re stuck.
Turns out AI platform adoption works the same way.
Your team wants to build everything on ChatGPT’s platform. Integrate deeply, make it central to your product, bet the roadmap on it. The distribution looks incredible—millions of users, perfect retention, obvious winner. Why wouldn’t you go all-in?
Because you’ve watched this movie before. You saw colleagues build Facebook apps in 2008, ride the wave, then get destroyed when Facebook closed organic distribution and absorbed the best use cases into first-party features. You watched companies build entire businesses on SEO, then Google suppressed organic results for ads and their own products. You saw iOS developers thrive, then get crushed by the 30% tax.
Younger managers see the opportunity and bet everything. They haven’t watched enough cycles to know the pattern. You have. You know platforms always close for monetization—it’s not evil, it’s capitalism and competitive dynamics. They have to keep growing, prevent their own disruption.
The insight isn’t “don’t play the game”—it’s a prisoner’s dilemma, you have to play. The insight is: ride the wave early, extract the distribution, but anticipate the closure and sequence away before it happens. Don’t build your entire business on someone else’s platform.
That judgment—knowing when to ride platforms versus when to diversify—comes from watching the full arc multiple times. Facebook, Google, Apple, LinkedIn all followed the same playbook. You know it’s coming.
Context
Brian Balfour founded Reforge and was VP of Growth at HubSpot. He lived through the Facebook platform boom with social gaming, watched Google SEO evolve over decades, observed Apple App Store dynamics, and recently saw LinkedIn close organic distribution twice (company pages, then personal profiles).
The four-stage platform cycle (conditions met → moat identified → platform opens → platform closes) repeats predictably. For experienced leaders evaluating AI platform bets, this isn’t theory—you’ve watched mobile, cloud, social, and SEO platforms all follow this exact pattern.
That pattern recognition is how you know to extract value without becoming dependent.