Contribution
A name and a three-stage schema for a pattern in platform businesses, and a policy conclusion drawn from it: since platforms will decay, protect users’ ability to leave rather than the platforms. It is an essay, not a study.
Key insight
A platform that can cheaply change who gets the value passing through it will spend surplus on whichever side it needs to attract and then take it back once that side is locked in. How far it can go is set by how costly leaving is, so the decisive variable is the cost of exit, not the platform’s intentions.
Assumptions
- Platforms are two-sided markets that hold each side “hostage to the other”.
- Reallocating value is cheap: “It’s just too easy to turn the enshittification dial up to eleven.”
- Firms’ internal actors prefer quick returns: product managers, executives and shareholders “are in a race to see who can eat their seed-corn first”.
- Users leave once abuse passes a breaking point, and the point rises with switching costs.
Key results
- The lifecycle, stated in the opening paragraph and restated as “surpluses are first directed to users; then, once they’re locked in, surpluses go to suppliers; then once they’re locked in, the surplus is handed to shareholders”.
- Amazon: subsidised prices and search, DRM and Prime; then generous terms for sellers; then fees and paid placement (the author’s figures, unchecked: 45% and more in fees, the first screens of a search half ads).
- Facebook: friends’ posts; then publishers’ traffic; then throttling publishers unless they pay to boost; then more ads and higher prices.
- TikTok: the Forbes report that staff “heat” chosen creators’ videos, read as surplus handed to creators to commit them, to be withdrawn later. Doctorow notes the limit: “there’s only so much enshittification they can visit upon their users’ feeds”.
- Interoperability: blocking clients and APIs makes leaving harder and so raises the abuse users will bear.
- Equilibrium: incentives push results “to the precise threshold at which users are nearly pissed off enough to leave, but not quite”, an equilibrium that is “fragile” to competitors, scandals and “a worker uprising”.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Platforms decay in three stages, good to users, then to business customers, then to themselves | weak | cases (Amazon, Facebook, Twitter, TikTok), told by the author |
| C2 | The cause is cheap reallocation of value in a two-sided market | weak | asserted; no comparison with platforms that did not decay |
| C3 | Lock-in and blocked interoperability raise the abuse users will tolerate | moderate | a coherent switching-cost argument, illustrated, not measured |
| C4 | Monopoly and acquisitions are why it has lasted | weak | asserted, with the Instagram purchase as example |
| C5 | TikTok will withdraw creators’ boosted reach once they are committed | weak | a prediction from one press report |
| C6 | Policy should secure end-to-end and freedom of exit rather than preserve platforms | — | a normative conclusion, argued |
Concepts
- enshittification — the three-stage lifecycle; “I call this enshittification”.
- two-sided market — a platform between buyers and sellers, “holding each hostage to the other”.
- end-to-end principle — “willing speakers’ messages would be delivered to willing listeners’ end-points as quickly and reliably as they could be”.
- freedom of exit — “the right to leave a sinking platform while continuing to stay connected to the communities that you left behind”, keeping media and data.
Connections
The DEF CON 32 talk (LIT-033, read in NOTE-041) restates the lifecycle a year and a half later and organises the checks on it as four forces: competition, regulation, interoperability and tech-worker power. This essay already has three of them in some form, competitors, interoperable clients and “a worker uprising”, but not regulation as a check; here lawmakers appear only as the source of the remedy. The 2025 book (LIT-753) is unread. The essay draws on the Netheads’ end-to-end tradition and on Page and Brin’s 1998 remark that advertising-funded search will be biased toward advertisers.
Bearing on the record
Source of THEORY-150. For the owner’s essay:
- §6.4 (alien interests). This is the essay’s cited source. Doctorow’s platform comes to pursue an interest, its shareholders’ returns, that is served by working against the users and businesses it is made of and for: “every time Tiktok shows you a video you asked to see, it loses a chance to show you a video it wants you to see”.
- §6.3 (extraction from above). The record reads the third stage as an instance of extraction from above: the platform, the larger unit, takes value from the users and businesses inside it. Doctorow does not use levels.
- §5.4 (protective lock-in). Doctorow’s lock-in is not protective; it is built so that abuse can follow. It shares the essay’s point that a regime can create the conditions that keep people in it.
Limitations
The evidence is anecdote, linked journalism and the author’s own experience, with no measure and no counter-cases: no platform that stayed good to its users is considered. “Seemingly inevitable” is hedged in the same essay by the need for “measured amounts”. The scope is platform capitalism, not organizations in general.
Open questions
Whether the lifecycle holds when switching costs are measured, and whether platforms with low lock-in decay less; whether the stages are a sequence or a description of a platform’s choices at any one time.