The Enshittification of the Internet: A Reverse-Engineering Guide

Enshittification is the process by which digital platforms systematically degrade user experience to maximize profit. The term, coined by Cory Doctorow, describes a three-stage cycle: platforms first attract users with value, then extract value from users to attract business customers, then extract value from everyone until the platform collapses or becomes a zombie institution sustained only by lock-in. This guide provides a reverse-engineering framework for understanding the mechanics of enshittification and practical strategies for resisting it.

If you've ever felt that a service you once loved has gotten worse — more ads, fewer features, higher prices, more dark patterns — you're not imagining things. You're experiencing enshittification. And understanding the pattern is the first step toward breaking free of it.

The Enshittification Cycle

Every platform that enshittifies follows the same three-stage trajectory. The stages are not accidental — they are the logical consequence of the venture-capital-funded platform business model. Understanding each stage allows you to identify where any given platform currently sits in the cycle, and to make informed decisions about whether to continue using it.

Stage 1: User Acquisition

In the first stage, the platform is generous. It subsidizes services, offers excellent user experience, supports interoperability, and generally behaves as if its primary mission is to serve users. This is the "too good to be true" phase, and it is, in fact, too good to be true.

During this stage, the platform is burning investor money to build a user base. The economics don't work — the service costs more to provide than it generates in revenue. But that's the plan: grow fast, achieve network effects, make the platform indispensable, and then move to Stage 2.

Examples of Stage 1 behavior:

• Google Search (early 2000s): Clean interface, no ads above the fold, genuinely useful results. The best search engine by a wide margin, offered completely free.
• Facebook (2004-2012): Chronological feed, no ads, easy photo sharing, genuine social connection. It felt like a public utility for staying in touch with friends.
• Amazon (late 1990s-2000s): Below-cost pricing, free shipping, excellent customer service, generous return policies. The company lost money for years to build its customer base.
• Spotify (2008-2015): Vast music library, clean interface, reasonable free tier, fair artist payments (relatively speaking). Music streaming felt like magic.
• Uber (2012-2017): Rides cheaper than taxis, fast pickup times, easy payment, friendly drivers earning good money. Subsidized entirely by venture capital.

Stage 2: Business Pivot

Once the platform has achieved sufficient scale and lock-in, it begins extracting value from users to attract business customers (advertisers, merchants, content creators who pay for reach). The user experience degrades, but not enough to trigger mass exodus — the lock-in from Stage 1 keeps people in place.

Stage 2 tactics include:

• Advertising injection: Ads appear in feeds, search results, and interfaces that were previously ad-free. The ads are initially labeled clearly; over time, the distinction between organic and paid content blurs.
• Data harvesting intensification: The platform collects more data, shares it more broadly, and uses it for increasingly invasive targeting. Privacy policies get longer and less comprehensible.
• Algorithmic manipulation: Chronological feeds are replaced by algorithmic ones that prioritize "engagement" (content that provokes strong emotional reactions) over relevance or quality. Users see what the algorithm decides, not what they chose.
• Feature removal or degradation: Free features become paid. API access is restricted or eliminated. Export tools are buried or disabled. The platform actively makes it harder to leave.
• Creator squeeze: Content creators who built audiences on the platform find their organic reach throttled. To reach the audience they already built, they must now pay for promotion.

Stage 3: Terminal Extraction

In the final stage, the platform extracts maximum value from everyone — users, business customers, and creators alike. Quality collapses. The service becomes a shell of its former self, sustained only by the switching costs and network effects that make leaving painful.

Terminal extraction looks like:

• Search results full of ads and SEO spam: Google's first page is now ads, AI-generated summaries, "People also ask" boxes, and featured snippets — the actual organic results are pushed below the fold.
• Social feeds dominated by recommended content from strangers: Instagram and Facebook show you more content from accounts you don't follow than from accounts you do. Your feed is no longer yours.
• Marketplace flooded with counterfeits and paid placements: Amazon's search results are a pay-to-play jungle. Sponsored products, fake reviews, and counterfeit goods make it nearly impossible to find what you're actually looking for.
• API shutdown: Twitter/X killed third-party clients, Reddit priced out third-party apps, and Google has systematically deprecated APIs that enabled independent innovation on their platforms.
• Price increases with declining quality: Streaming services raise prices while reducing content libraries and introducing ad tiers. Subscription fatigue is a direct consequence of terminal extraction across multiple platforms simultaneously.

Why Enshittification Happens

Enshittification is not a bug — it is a feature of the venture-capital-funded platform business model. Understanding the structural causes reveals why individual companies aren't really the problem; the incentive structure is.

• Venture capital growth mandates: VC-funded companies must grow at rates that justify their valuations. A company valued at $10 billion must generate returns that justify that valuation, which means extracting far more value than a sustainable business would. The growth imperative makes enshittification mathematically inevitable once organic growth slows.
• Network effects as lock-in: The value of a social network, marketplace, or communication platform depends on how many other people use it. Once a platform achieves critical mass, leaving means losing access to your network — your friends, your customers, your audience. This lock-in gives the platform enormous power to degrade quality without losing users.
• The advertising business model: When the customer (the advertiser) and the user (you) are different entities, the platform's incentives diverge from yours. Your attention is the product being sold. The platform is optimizing for the advertiser's satisfaction, not yours. Any improvement to your experience that reduces ad revenue will eventually be reversed.
• Regulatory capture: Large platforms invest heavily in lobbying, ensuring that regulations either don't exist or are written in ways that entrench incumbents and raise barriers to entry for competitors.
• Switching costs and data portability barriers: Platforms deliberately make it difficult to export your data, transfer your social graph, or move your content to a competitor. The harder it is to leave, the more abuse you'll tolerate.

The Reverse-Engineering Toolkit

Understanding enshittification is necessary but not sufficient. The goal is to develop practical strategies for resisting it — or at least for minimizing its impact on your life. Here is a four-part toolkit.

Identify the Stage

For every platform you use regularly, determine where it sits in the enshittification cycle. This isn't always obvious — platforms don't announce their transition from Stage 1 to Stage 2. But there are reliable signals:

• Are there more ads than there were a year ago?
• Has the algorithm changed what you see without your consent?
• Have features been removed, paywalled, or degraded?
• Has the platform restricted API access or third-party integrations?
• Does the platform make it harder to export your data than it used to?
• Has the platform started showing you content from accounts you don't follow?

If you answered yes to three or more of these, the platform is likely in Stage 2 or early Stage 3. Plan your exit strategy accordingly.

Reduce Lock-in

The platform's power over you is proportional to your lock-in. Every step you take to reduce lock-in increases your freedom to leave when the enshittification becomes intolerable.

• Export your data regularly. Most platforms offer data export (often buried in settings). Use it. Keep local copies of your photos, posts, messages, and contacts.
• Use interoperable formats. Write in Markdown, not in a proprietary editor. Store files in open formats (PDF, PNG, MP3, CSV) rather than platform-specific ones.
• Maintain presence on alternatives. Even if you primarily use one platform, keep an account on an alternative. When the time comes to switch, you won't be starting from zero.
• Build your audience on owned property. If you create content, your home base should be a website or newsletter you control — not a social media profile that can be throttled, banned, or enshittified at any time.

Use Alternatives

For nearly every enshittified platform, a better alternative exists — often open-source, federated, or community-owned. The tradeoff is usually convenience and network size for quality and autonomy.

• Search: Kagi (paid, no ads, excellent quality), DuckDuckGo (free, private), SearXNG (selfhosted metasearch).
• Social media: Mastodon/Fediverse (federated, no algorithm), Bluesky (decentralized protocol), or simply RSS + blogs.
• Email: Fastmail, Proton Mail, or selfhosted (instead of Gmail, which reads your email to target ads).
• Cloud storage: Syncthing + local NAS (instead of Google Drive/Dropbox).
• Maps: OpenStreetMap + Organic Maps (instead of Google Maps, which tracks your location history).
• Video: PeerTube (federated), Nebula (creator-owned), or NewPipe/Invidious as YouTube front-ends that strip tracking and ads.
• Messaging: Signal (encrypted, non-profit), Matrix/Element (federated, encrypted), or XMPP (the original open messaging protocol).

Build Community Infrastructure

Individual action is important but insufficient. The deepest resistance to enshittification comes from building and supporting alternatives that are structurally resistant to the cycle — because they have different ownership models and incentive structures.

• Support cooperatively owned platforms where users are owners and the incentive is to serve members rather than extract from them.
• Contribute to open-source projects that provide alternatives to enshittified services. Code, documentation, bug reports, and financial contributions all help.
• Advocate for local-first software — applications that store data on your device and sync peer-to-peer, rather than requiring a central server that can be enshittified.
• Support legislation that mandates interoperability, data portability, and the right to exit platforms with your data and social graph intact.

Beyond Individual Action

Personal choices matter, but enshittification is a structural problem that requires structural solutions. Several policy and legal frameworks are emerging to address it:

• The EU Digital Markets Act (DMA) designates large platforms as "gatekeepers" and imposes interoperability requirements, bans self-preferencing, and mandates data portability. It is the most ambitious regulatory response to platform power in history.
• Right to repair legislation is expanding from physical hardware to digital services. The principle that you should be able to fix, modify, and understand the tools you use is being extended to software and platforms.
• Adversarial interoperability — the practice of building tools that work with a platform without its permission (ad blockers, alternative clients, data scrapers) — is a powerful form of resistance. Doctorow argues that legalizing and protecting adversarial interoperability would do more to fight enshittification than any other single policy.
• Community-owned infrastructure — from cooperatively owned broadband to community-run social media instances — provides alternatives that are structurally resistant to enshittification because they don't have shareholders demanding infinite growth.

The enshittification cycle is not inevitable. It is a consequence of specific ownership structures, business models, and regulatory environments. Change those, and you change the outcome. In the meantime, understanding the cycle gives you the power to make informed choices about which platforms deserve your time, your data, and your trust — and which ones have forfeited all three.

Further Reading

• Cory Doctorow, "Tiktok's Enshittification" (2023) — The original essay that named and defined the pattern.
• Cory Doctorow, The Internet Con: How to Seize the Means of Computation (2023) — A book-length treatment of enshittification and the case for adversarial interoperability.
• What Is Cognitive Sovereignty? — A complementary framework from 421.news on protecting mental autonomy from platform manipulation.
• Low Tech, High Life — A practical framework for choosing simpler, more intentional tools over enshittified platforms.
• Shoshana Zuboff, The Age of Surveillance Capitalism (2019) — The definitive analysis of how platforms extract and monetize behavioral data.
• Tim Wu, The Attention Merchants (2016) — A historical account of how attention has been captured and sold, from newspapers to social media.