TikTok and airlines have something in common with your search engine, your grocery app, and (increasingly) your car: They start out great, lock you in, and then quietly get worse while you keep using them. That very familiar decline now has a catchy name: \u201censhittification.\u201d<\/p>\n
Cory Doctorow has been writing about this for decades as a journalist, activist with the Electronic Frontier Foundation, and science-fiction author. His new book, Enshittification: Why Everything Suddenly Got Worse and What to Do About It<\/a>, is a field guide to how platforms decay, why they get away with it, and what it will take to reverse course.<\/p>\n I invited Doctorow onto The Gray Area to map the lifecycle of a platform, explain the policy choices that made today\u2019s tech feudalism possible, and outline the structural fixes that could make the internet (and the economy around it) less extractive and more humane.<\/p>\n As always, there\u2019s much more in the full podcast, so listen and follow The Gray Area<\/a> on Apple Podcasts<\/a>, Spotify<\/a>, Pandora<\/a>, or wherever you find podcasts. New episodes drop every Monday.<\/p>\n This interview has been edited for length and clarity.<\/p>\n What is \u201censhittification\u201d? Give me the cleanest definition.<\/strong><\/p>\n At the descriptive level, it\u2019s a pattern in how platforms go bad. First, they\u2019re great to end users. Then they find ways to lock those users in \u2014 switching costs, network effects, contracts, DRM<\/a> \u2014 and once users are stuck, the company makes the product worse for them to extract more value. Next, they use that surplus to woo business customers (advertisers, sellers, creators), lock them in, and start making the product worse for the business side too. Eventually, everyone is trapped and the platform turns into a pile of crap. You can see this in places as different as Google, Facebook, Uber, and Amazon.<\/p>\n The more interesting question is: Why now? Greed isn\u2019t new. Venture capital isn\u2019t new. What changed were the constraints on firms, especially the degree of competition and the legal environment that lets platforms \u201ctwiddle\u201d the experience for each user, while blocking users and rivals from pushing back.<\/p>\n When did the \u201cgreat enshittification\u201d begin? Can you put a stake in the ground?<\/strong><\/p>\n There are obvious post-enshittification moments. In 2019, for example, the Google antitrust case records show an internal clash: Google had 90 percent market share in search, growth had stalled, and an executive pitched a strategy to make search worse so users would have to run multiple queries and see more ads. That\u2019s enshittification in a nutshell \u2014 and we all kept using Google anyway.<\/p>\n But it\u2019s not a single date. The defining feature isn\u2019t \u201cthings got worse\u201d \u2014 it\u2019s \u201cthings got worse and we stayed.\u201d The preconditions \u2014 consolidation, policy choices, and legal shields \u2014 built up over years.<\/p>\n Why do we keep using products after they get worse? Why not just leave?<\/strong><\/p>\n \u201cPlatforms can constantly tweak what you see and what you pay, while users and independent developers are barred from looking under the hood or restoring balance.\u201d<\/p>\n The galaxy-brain answer blames consumers for \u201cshopping wrong,\u201d or says, \u201cIf you\u2019re not paying for the product, you are the product.\u201d That\u2019s not it. The real story is diminished constraint [on market consolidation]. It\u2019s hard to take your business elsewhere when there is nowhere else to go.<\/p>\n That was policy-driven. For decades, antitrust enforcers embraced the Chicago School idea that monopolies<\/a> are efficient and should be punished only if prices go up for consumers \u2014 not when a company buys its most dangerous rival. Facebook\u2019s purchase of Instagram is the classic example. Mark Zuckerberg literally wrote that people were leaving Facebook for Instagram and that buying Instagram would keep them as Facebook users even if they never touched Facebook again. That\u2019s an antitrust admission in plain English. And yet the Obama administration waved it through, just as the Bush and Trump administrations green-lit their own waves of consolidation.<\/p>\n You emphasize that this corrosion is especially intense on <\/strong>platforms<\/strong>. Why are platforms such fertile soil for enshittification?<\/strong><\/p>\n Because platforms sit between two groups that need each other: end users and business customers. That intermediation is useful, and most of us don\u2019t want to process payments or write our own mail servers just to publish a newsletter. But digital platforms have a unique superpower: They can change the business logic on a per-user, per-interaction basis. I call this \u201ctwiddling.\u201d<\/p>\n Combine that with legal shields, and you get a one-way ratchet. Anti-circumvention laws (think DMCA \u00a71201<\/a>) make it illegal to reverse-engineer or \u201cmod\u201d an app or device to inspect, verify, or override what it\u2019s doing \u2014 even for lawful purposes like price comparison or accessibility. Platforms can constantly tweak what you see and what you pay, while users and independent developers are barred from looking under the hood or restoring balance.<\/p>\n Is this just a platform story, or is <\/strong>everything<\/strong> enshittifying?<\/strong><\/p>\n It\u2019s spreading. When Jeff Bezos wants to change Amazon Fresh prices, he moves a slider. When he wants to change a Whole Foods shelf, he needs an army with price guns \u2014 unless the shelf labels are e-ink tags, which are rolling out everywhere. Once a sector is digitized, it\u2019s platformized \u2014 and the twiddling follows. That\u2019s how you get \u201cdynamic\u201d prices for fast food or \u201csurge\u201d pricing at the drive-thru. Even when companies walk back the PR, the trial balloon shows what\u2019s technologically trivial and what\u2019s hard for us to detect without the legal right to inspect the software.<\/p>\n Let\u2019s ground this in a few case studies. Facebook feels like the poster child. How did it enshittify?<\/strong><\/p>\n Early Facebook wooed users by promising reverse-chronological feeds from the people you chose \u2014 no surveillance, no \u201cboosted\u201d junk. Users piled in. Then the collective-action trap kicked in: you love your friends, you hate the platform, but you can\u2019t coordinate a mass move. The larger Facebook gets, the harder it is to leave.<\/p>\n Once users were locked, Facebook twiddled to please business customers: \u201cGive us a little money and we\u2019ll target your ad to exactly who you want.\u201d \u201cPublishers, put teasers on Facebook and we\u2019ll stuff them into eyeballs that didn\u2019t ask to see them; you\u2019ll get a free traffic funnel.\u201d That locked the businesses in, too \u2014 a monopsony, where one buyer dominates sellers.<\/p>\n Then came the second squeeze: Ad targeting fidelity declined, fraud rose, and prices went up. Procter & Gamble famously cut $200 million<\/a> in \u201cprogrammatic\u201d (surveillance) ads and saw no drop in sales, because so much of that spend was disappearing into the fraud hole. Meanwhile, users\u2019 feeds filled with paid content until what they asked to see was homeopathic residue. Everybody hated it; almost nobody left. That brittle equilibrium \u2014 \u201cI hate this place but can\u2019t stop visiting\u201d \u2014 is where platforms mint money.<\/p>\n How did Amazon enshittify?<\/strong><\/p>\n Start with the demand side. In a consumption-driven economy with rising inequality, a shrinking share of households buys a disproportionate share of stuff, and most of those households have Prime. If you sell goods, you have to be on Amazon, and you have to pay to be found on Amazon.<\/p>\n Amazon\u2019s ad business \u2014 it was about $32B when I wrote the book; it\u2019s now north of $50B \u2014 isn\u2019t \u201cadvertising\u201d in the classic sense. It\u2019s payola: Pay to be placed at the top of search results. That means the top result is often not the best price or best quality; it\u2019s who paid the most. On average, the first result is about 29 percent more expensive than the actual best deal; the entire top row is about 25 percent pricier; the true best buy is often buried around result 17 \u2014 into page two. Sellers game this by, say, selling four-packs instead of 10-packs so they look cheapest on a per-item sort, while the unit price is worse. Amazon knows unit sizes. It could let you sort by unit price. It doesn\u2019t.<\/p>\n \u201cStructural problems need structural fixes.\u201d<\/p>\n Then there\u2019s \u201cmost-favored nation\u201d clauses: If you [as a goods-selling business] raise prices on Amazon to recoup Amazon\u2019s junk fees, you have to raise them everywhere else, including your own site. That\u2019s how Amazon\u2019s economics spread off-platform. And on the lock-in side, digital goods (Kindle books, Audible audiobooks, video) are wrapped in DRM; you [as a customer] leave Amazon, you lose your library. Prime pre-pays your shipping, and subscription mechanics (one credit per month) nudge you to keep using the store you\u2019ve already funded. Exclusivity deals make rivals\u2019 catalogs worse by design.<\/p>\n What about the \u201cyou don\u2019t <\/strong>have<\/strong> to use Amazon\u201d objection, because you can always just shop somewhere else?<\/strong><\/p>\n There are switching costs and there\u2019s the scorched earth enshittification leaves behind. Uber is instructive: It subsidized rides to kill off taxis, then raised prices. In cities like mine, there might be a dozen cabs left; buses cut service because \u201ceveryone takes Uber\u201d; on a 115-degree day, your \u201cchoice\u201d is an Uber Black at $96 for a mile and a half or\u2026walking with a suitcase. The alternatives exist \u2014 they just suck more now.<\/p>\n That\u2019s why I tell people: Don\u2019t moralize your shopping cart. Zephyr Teachout makes this point well: If you skip a labor rally at the warehouse because you\u2019re driving around looking for artisanal poster markers, so your sign won\u2019t be \u201ctainted\u201d by Amazon, Jeff Bezos wins. You are not an ambulatory wallet. Structural problems need structural fixes.<\/p>\n Your book points to four buckets: antitrust, regulation with teeth, interoperability (the right to move between services and jailbreak your stuff), and worker power. Which is most achievable right now?<\/strong><\/p>\n In the US, the antitrust coalition that briefly united Ted Cruz and Elizabeth Warren has mostly evaporated; the vibe is transactional. I\u2019m more optimistic about foreign antitrust \u2014 the EU, UK, and others are motivated to reduce their dependence on American platforms, especially when US firms admit they can\u2019t guarantee EU state secrets won\u2019t be reachable under secret US orders. That geopolitical incentive produces muscular cases and new rules (see the EU\u2019s DMA\/DSA<\/a>).<\/p>\n Worker power is also ripening. Tech workers are realizing they\u2019re workers \u2014 not \u201cfounders in waiting.\u201d The average Silicon Valley engineer generates about $1 million in revenue; now there\u2019ve been hundreds of thousands of layoffs, and management is dangling AI as a way to make you feel replaceable. When bosses aren\u2019t afraid of you, you get Foxconn-style treatment. Unions are the answer. Yes, the administration has kneecapped the NLRB, but that misunderstands the history: Unions didn\u2019t arise because law permitted them; law permitted unions because workers organized anyway. If you\u2019ve fired the referees, there are no rules left \u2014 and solidarity across sectors becomes even more powerful.<\/p>\n