Most people can name an app that used to be better. Maybe it was a search engine that put the answer right at the top of the page, a video service that felt like a bargain, or a social feed that mostly showed friends and family instead of strangers, bizarre suggestions and sponsored posts. Few of us can point to the day things changed, but like the frogs in the boiling pot, we keep using services that seem to be progressing backwards. First comes a product that is simple, generous and often cheap. Then come small changes that are easy to overlook, like one more banner, a functionality that moves behind a paywall, or a setting that quietly resets itself. Eventually the feature that won you over is hard to find. What makes this more than a matter of taste is how consistent the pattern looks. A shopping site, a ride service, a photo network and a streaming library have little in common on paper, yet their users describe a strikingly similar slide. Sellers and advertisers, the very customers these companies court, often voice the same complaints. Economists and authors have started to examine this deterioration more closely, and in early 2024 a society of linguists chose a word describing it as Word of the Year, ahead of several contenders about artificial intelligence. So what is it really and how does it happen? aleksandarlittlewolf / magnific (not the actual photo) The word is enshittification, and it is exactly as blunt as it sounds. The American Dialect Society made it its 2023 Word of the Year, defining it as the worsening of a digital platform through a reduction in quality of service. Writer and activist Cory Doctorow used it in a January 2023 essay about TikTok that Wired republished within days, describing how an online platform can slide from serving its users to just extracting value from them. Doctorow turned the idea into a 2025 book, and other writers like Tim Wu expanded on it. We'll come back to both of them. To see how a product ends up on that path, it helps to look at how most apps and online services begin. A young company rarely starts by earning money. Uber famously didn’t really turn a profit before 2023. A common route is to launch something simple, raise funds from venture capital investors who accept years of losses in exchange for a shot at an enormous payoff, and use that money to keep prices low or at zero so the audience grows quickly. For much of the 2010s and the early pandemic years, interest rates sat close to zero, which made it easier for tech companies to raise money and keep growing. Solen Feyissa / Unsplash (not the actual photo) Then central banks raised rates in 2022. This meant a shift away from "growth at all costs" toward a clear route to profit, and analysts have described the end of subsidized growth for a generation of startups built in the cheap money era. Once a service has a large audience, the central question changes from how to attract users to how to earn from them, whether through ads, subscriptions, fees or data. As a user, this is often bad news. The product tends to pick up investor reporting, revenue targets and managers measured on business results, so new features are judged by what they add for the company with usability as a second thought. Doctorow’s model describes what often follows in three stages. Initially, the platform is a good deal for its users, often using investor money to offer better prices than anyone else. In the second, once people are hooked, it shifts value toward business customers such as advertisers and sellers, at some cost to users. In the third, it takes more of the value for itself and squeezes both groups. The glue holding this together is user familiarity. When your friends, purchases or customers are all in one place, leaving is costly, so a platform can change the terms without losing many people. On social networks, users supply much of the “lock in“ themselves, because leaving means persuading friends and family to move too. This can be just as true for the things we purchase online, audiobooks sold with digital rights management that tie them to a single seller. deneyterrio / flickr Some companies take offense at the idea that they’ve enshittified their products. During the 2025 antitrust trial against Meta, the company presented figures showing that people spend about 17 percent of their time on Facebook and 7 percent on Instagram viewing posts from friends, with much of the rest going to content from accounts they do not follow. Meta presents this as a reflection of how people now choose to use its apps, while Doctorow argues that feeds were rebuilt around advertisers. Brett Jordan / Unsplash (not the actual photo) At Amazon, one analysis of sellers’ financial statements found that the company’s cut of a typical seller’s revenue passed 50 percent in 2022, up from about 40 percent five years earlier, as fulfillment fees rose and advertising became hard to avoid. The same analysis notes that those services would cost money elsewhere too, and Amazon has called Doctorow’s theory about the company incorrect, saying customer value has kept improving. Streaming offers a gentler case. Netflix added a cheaper tier with ads, after years of bragging that they don’t run them and began charging for shared passwords and raised the price of its premium plan without ads to $22.99 in 2023, yet it still added millions of subscribers. A fair question is why so many platforms seem to slide around the same time. Doctorow notes that Microsoft and Apple are decades older than Facebook, so if decline were simply a matter of age it would have reached them at very different times. His explanation is that several constraints weakened together: competition thinned after consolidation, regulators moved slowly, and workers lost leverage after layoffs. He also points to laws such as the US Digital Millennium Copyright Act, which can make it a crime to bypass digital locks on software, blocking the kind of outside fix that might pressure a company to behave. Printer cartridges are another example, where a thin layer of software can make cheaper rival cartridges legally risky to sell. Solen Feyissa / Unsplash (not the actual photo) Not everyone accepts the full theory. An economic analysis by Paul Belleflamme, published in October 2025, finds support for some parts of the thesis and challenges others. It notes that dominance does not necessarily doom a platform, and that decay can also come from fierce competition among rivals instead of too little of it. Proposals for dealing with the problem tend to fall into a few families, and most try to make it easier for people to leave or for rivals to compete. The European Union’s Digital Markets Act is probably one of the more cohesive attempts. It places obligations on the largest designated platforms around data portability, interoperability and self preferencing, and the European Parliament called for rigorous enforcement of them in April 2026. Apple, one of its critics, argued in 2025 that the law creates security risks and delays features, while Brussels maintains that it exists to ensure fair competition. Sara Oliveira / Unsplash (not the actual photo) Another approach targets the pain points users feel most directly. In the United States, a federal rule requiring subscriptions to be as easy to cancel as to start was vacated by an appeals court in 2025 on procedural grounds, and the Federal Trade Commission opened a fresh rulemaking in March 2026 while continuing case by case enforcement under existing law. A third idea is to give outsiders legal room to act. Doctorow argues for revisiting laws that criminalize bypassing digital locks, so rivals and startups could build the fixes a platform will not, while Wu favors neutrality rules and limits on tech giants entering new markets. None of these is a guaranteed cure. Paul Belleflamme’s analysis notes that platform decay can also arise from intense competition among rivals, so more competition is not automatically the answer. Even so, the proposals share a premise, which is that platform behavior responds to incentives and rules, and just maybe a company can be convinced to not flush it's app down the drain.
