Enshittification Was Engineered Before It Had a Name.

In this article:

• What Moglen Identified Before Doctorow Named It
• How the Surveillance Business Model Produces Enshittification
• Why the Platform Cannot Self Correct Once Extraction Begins
• What Enshittification Looks Like From Inside the SaaS Stack
• Frequently Asked Questions
• What is enshittification in SaaS platforms?
• What is the connection between surveillance capitalism and enshittification?
• Why does the investor funded SaaS model trend toward enshittification?
• References

Cory Doctorow named enshittification in a 2022 essay about TikTok, and the term spread because it named something people already recognized. The thing it named had been happening for decades before the word existed. Eben Moglen described the mechanism in a 2010 lecture at the Internet Society of New York without using the term. What he described was the same sequence, derived from the same architectural logic, producing the same outcome. The mechanism was diagnosed before it had a name. The name arrived after the evidence was already overwhelming.

In 2010, Moglen stood in front of the Internet Society of New York and described what Facebook was doing with unusual precision. The platform offered free web hosting, PHP utilities, and social features in exchange for behavioral data collected continuously without meaningful user understanding of what was being collected or how it would be used. The value exchange was explicit in the terms of service and invisible in the experience: users received a social platform; the platform received a surveillance feed of user behavior.

Moglen's framing was architectural rather than moral. He was not primarily arguing that Facebook was behaving badly. He was arguing that the architecture Facebook was built on made its behavior structurally inevitable. "The business model that supports them is misuse," he said of Facebook and similar platforms. "There isn't any other business model for them." This was not a description of something that had gone wrong. It was a description of something that had gone according to plan.

What Moglen diagnosed was the engine beneath enshittification: a platform that offers genuine value in its early phase because the genuine value is what attracts the users whose behavioral data makes the business model work. The free social platform is not a gift that later gets monetized. The social platform is the mechanism for producing the behavioral data asset, and monetizing that asset is the plan from the beginning. The free service is the cost of acquiring the data. The users are not the customers. They are the product being refined for sale.

Doctorow's enshittification framework made explicit what Moglen's architectural argument implied: the value phase is phase one, designed to end. It ends when the lock in achieved through genuine value creation is sufficient to support extraction. At that point the value degrades, the extraction increases, and the user has already built enough into the platform that leaving costs more than absorbing the worse deal. Moglen described the same sequence in 2010. Doctorow named it in 2022. The platforms ran it continuously in the twelve years between.

The significance of the gap between Moglen's diagnosis and Doctorow's naming is not historical. It is structural. Enshittification was not discovered after platforms went bad. It was built into the architecture before the platforms launched. The decay is not a product of failure. It is a product of the model working exactly as designed.

Shoshana Zuboff's concept of surveillance capitalism provides the economic mechanism that connects Moglen's architectural diagnosis to Doctorow's enshittification sequence. Surveillance capitalism is not simply data collection. It is the extraction of behavioral surplus: the data generated by human activity that exceeds what is needed to improve the product for the user, which is then used to build predictive models and sold to third parties, typically advertisers targeting future behavior.

The business model creates a specific and predictable incentive structure. The platform needs behavioral data to sell. More behavioral data means better predictive models means higher advertising revenue. To maximize behavioral data, the platform has an incentive to maximize user engagement. To maximize engagement, the platform invests in making the experience more compelling and more time consuming, not necessarily more useful or more aligned with what users actually want. The platform that optimizes for engagement metrics rather than user value is already in the early degradation phase of enshittification even if its subscription count is growing and its reported metrics look healthy.

This is where the surveillance capitalism model and the enshittification sequence converge. The platform that generates behavioral surplus to fund its growth is already extracting more than it is delivering in value. The user experience may still be positive. The extraction is happening in the background: in the behavioral data the user does not know is being collected, the predictive model being built about them without their knowledge, the advertising sold against their predicted behavior. Enshittification in its early phase is invisible to users precisely because the product is still good. The extraction is happening through a different channel than the product quality the user is evaluating.

Zuboff described behavioral surplus as data about people that they did not consent to produce and cannot access or contest. Moglen described the same thing in architectural terms as logs: "vast repositories of hierarchically organized data about people at the edges of the network that they do not control and unless they are experienced in the operation of servers will not understand the comprehensiveness of, will not understand the meaningfulness of, will not understand the aggregability of." Both were describing the same asset, accumulated through the same mechanism, for the same commercial purpose.

The structural argument for why enshittification is the model and not a deviation from it is developed in detail in Enshittification Is Not a Bug in SaaS. It Is the Model. The architectural basis that makes SaaS extraction compounding over time is documented in Cloudification Is Not a Technology Story. It Is a Power Story.

The enshittification sequence is structurally irreversible within the same ownership and incentive model, and the reason is not complicated. Enshittification begins when lock in is sufficient to support extraction. Reversing it would require reducing extraction, which would reduce returns to investors who funded the platform's growth on the expectation of being in the extraction phase. The investors did not fund the growth phase because they believed in the mission. They funded it because the growth phase was designed to produce lock in that would enable the extraction phase to be profitable.

There is no mechanism inside an investor funded platform for the decision to extract less. The decision might be made by an individual executive for strategic reasons and reversed by the board. It cannot be made structurally because no one inside the governance structure has both the authority and the incentive to make it last. The board answers to investors. The investors expect returns proportional to the risk they took funding the growth phase. The returns come from extraction. Reducing extraction is reducing returns, which is not a decision a board makes voluntarily while investors remain on the cap table.

Moglen saw this clearly in 2010: "It isn't technically innovative. It depends upon an architecture subject to misuse and the business model that supports it is misuse." His proposed corrective was not regulatory pressure or internal reform. It was a technical alternative that made the platform unnecessary. "When there is a competitor to all spying all the time, whether you like it or not, the competition's going to do real well." The corrective is not to reform the platform. It is to make the platform replaceable by infrastructure whose governance structure does not produce extraction.

The period between Moglen's lecture and Doctorow's naming of enshittification is twelve years of case studies confirming both the structural analysis and the absence of self correction. Reddit built the most valuable moderation community on the internet and then repriced its API to destroy the third party clients that made moderation sustainable, waited out the moderator revolt, and continued. Mailchimp built the small business email category, was acquired by Intuit, and restructured its pricing in ways that made the freelancers who built its reputation uneconomical to retain. Shopify evolved its fee structure in ways that made each additional year of use more expensive than the one before it. None of these corrected from inside.

For a business running on SaaS infrastructure, enshittification is an operational experience before it is a theoretical framework. The table below maps the three phases against documented examples from platforms that serve the independent business and agency market.

Each individual event in the extraction phase can be rationalized. Prices increase with inflation. Products evolve as companies mature. Support scales differently at different company sizes. The individual explanations are not false. What they explain away, when taken one at a time, is the pattern that only becomes visible when you stop taking them one at a time.

Varoufakis's concept of cloud capital adds the macroeconomic dimension to this operational experience. What the platform is extracting is not just subscription revenue. It is a portion of the value you produce using their infrastructure, captured through behavioral data generated by your use and through the pricing leverage their lock in creates. The subscription fee is rent. The behavioral surplus is a form of crop sharing. The business farms the platform's land; the platform collects a share of what grows there.

The three phase sequence applied to specific platforms with documented examples is detailed in Enshittification and the SaaS Decay Cycle. The argument for treating enshittification as a reason to restructure the infrastructure relationship rather than simply switch vendors is developed in BDS as Framework: Divesting From Your Extractive SaaS Stack.

Recognizing enshittification as a structural phenomenon rather than a quality complaint changes what the appropriate response is. A quality complaint is addressed by switching to a better product from a different vendor. A structural phenomenon is addressed by changing the infrastructure relationship itself. Enshittification does not mean the platform you currently use is uniquely bad. It means the ownership model governing any platform you depend on is eventually going to arrive at the extraction phase. The question is not whether your current tools will enshittify. The question is when, and whether the infrastructure relationship you are in will make it expensive to respond when they do.

Enshittification, as named by Cory Doctorow, is the decay sequence in which platforms build genuine value to attract users, then degrade that value to extract maximum revenue from users already locked in. In SaaS, enshittification arrives when switching costs are high enough that the vendor can raise prices, reduce quality, or change terms without losing customers, because leaving costs more than staying.

Surveillance capitalism, as Shoshana Zuboff described it, is the extraction of behavioral data from users beyond what is needed to improve the product. Enshittification is the product degradation that follows once lock in is established. The two are the same business model in sequence: first extract behavioral surplus to fund growth, then extract subscription revenue from the dependency that growth produces.

Investor funded SaaS requires a return that recurring subscription revenue alone rarely satisfies at exit. The path to exit through acquisition or IPO rewards growth metrics and lock in over product quality. Once lock in is achieved, the incentive to invest in product quality weakens and the incentive to increase extraction strengthens. The subscription model does not cause enshittification; the ownership and incentive structure of investor funded growth does.

Moglen, Eben. Freedom in the Cloud. Internet Society of New York, 2010.

Doctorow, Cory. Pluralistic: TikTok's enshittification. January 2023.

Zuboff, Shoshana. The Age of Surveillance Capitalism. PublicAffairs, 2019.

Varoufakis, Yanis. Technofeudalism: What Killed Capitalism. Bodley Head, 2023.