Do platforms inevitably decline through value extraction cycles?

Doctorow argues that platforms follow a predictable lifecycle, which he calls enshittification, in three phases: "first, they are good to their users; then they abuse their users to make things better for their business customers; finally, they abuse those business customers to claw back all the value for themselves." The surplus a platform creates moves in turn to users, then to suppliers once they are locked in, then to shareholders. He offers Amazon Marketplace, Facebook and Twitter as cases, and says the same lifecycle runs "from mobile app stores to Steam, from Facebook to Twitter."

The mechanism has two parts: the ease of changing how a platform allocates value, and the "two sided market" in which a platform "sits between buyers and sellers, holding each hostage to the other, raking off an ever-larger share of the value that passes between them." Facebook is his worked example. Once enough of a user's friends were on it, "it became effectively impossible to leave," and the publications and sellers that depended on those readers became dependent on the platform too, which let Facebook raise ad prices. The excerpt's figures, such as Marketplace sellers handing "45%+ of the sale price to Amazon in junk fees" and the first five screens of "cat bed" results being "50% ads," are Doctorow's own reported numbers. The excerpt gives no sample or method for either.

Set against the nearest notes, the lifecycle gives a structural reason feeds are not neutral. How do feed ranking weights shape what content gets produced? shows one weight changing what parties posted; Doctorow asks the broader version, "what if the logic shifts based on the platform's priorities?", and answers that it does, phase by phase. Creators are where this bites. Does AI content displace human influencers on social media? locates the threat to social proof in AI authorship. Doctorow locates creators' reach in platform allocation: his threads "used to routinely get hundreds of thousands or even millions of reads," and now get "hundreds, perhaps thousands." The two threats differ, but both leave a creator's audience dependent on decisions the creator does not control. The reported heating practice in Does TikTok use special boosts to inflate partner videos? is one account-level instance of the early-stage lure the lifecycle describes.

The excerpt does not establish the lifecycle's inevitability. Doctorow calls it "a seemingly inevitable consequence," but gives no comparison with platforms that did not decline, and his cases are illustrative rather than sampled. Several supporting figures are unsourced in the excerpt. It also says nothing about generative AI content, so extending the lifecycle to AI-written posts is a step the excerpt does not take. What the evidence supports is narrower: the lifecycle is a diagnostic for asking which side of a platform a recent change serves, not a timetable for when a platform will decline.