Enshittification: How Every App You Love Is Designed to Betray You

There's a word for it: Enshittification. The process by which a platform, service, or product that was genuinely good slowly, deliberately becomes worse. Extracting more value from users while delivering less of it until the product is a hollow, expensive, frustrating ghost of what it once was.

You see it everywhere. It's not an accident. It's the business model.

The Three-Stage Playbook

E.L Doctorow has a formulation that describes a consistent three-act structure. First, the platform is

good to users, subsidizes them, gives them genuine value, and wins their loyalty and habits.

Once users are locked in, the platform turns on them and starts extracting value: worse terms, higher prices, algorithmic friction, degraded service. The platform turns on the businesses that depend on it, squeezing suppliers, sellers, and workers for whatever margin is left. Then it collapses or pivots, leaving everyone worse off except the investors who cashed out at the top.

The fuel for stage one is almost always the same: venture capital or private equity money that lets the company operate at a loss to acquire users, buy market dominance, and destroy competitors. The company isn't profitable. It's buying your loyalty with investors' money. Once the competition is gone and you're dependent on the platform, the losses have to stop and you're the one who makes up the difference.

Uber: The Purest Example

Remember when Uber was genuinely cheaper than a taxi? That wasn't a business model. That was subsidized destruction. Uber lost billions of dollars annually for years, money funded by SoftBank and other investors to price taxis out of existence, make drivers dependent on the platform, and make riders habituated to app-based transport.

Once Uber had achieved dominance and competitors were dead or diminished, prices went up. Surge pricing became normalized. Driver pay got squeezed algorithmically. The service got worse. The app got more cluttered. Cancellation fees appeared.

Uber has still never consistently turned a profit on its core ride-hailing business. What it has done is capture a market and shift the cost of its original subsidies onto drivers and riders. The users got the product they were trained to expect. The investors got the exit.

The original gig economy pitch was freedom. And in the early days, it was partially true. Uber's founding drivers made good money because Uber was subsidizing the product to grow. DoorDash offered real earning opportunities when it was buying market share from restaurant delivery incumbents. TaskRabbit was a genuinely useful marketplace where skilled workers could find work on their terms.

But then, Enshittification of the workforce is what every gig platform looks like now. And Gig Economy 2.0, the new wave of AI-native platforms, has built stage two in from the beginning.

The Algorithmic Boss Has No Conscience

Across a growing number of platforms, human managers have been replaced entirely by automated systems. Your work is assigned by an algorithm. Your performance is rated by an algorithm. Your pay rate is set by an algorithm that adjusts in real time based on labor supply, your metrics, and variables you'll never be told about.

When you're deactivated the decision was made by an algorithm. The appeal process, if one exists, routes to a support queue staffed by underpaid contractors who cannot override system decisions.

Uber drivers report accounts being deactivated after a spike in low ratings, with no ability to contest individual ratings, no knowledge of which interactions triggered the threshold, and no recourse beyond a form email. Amazon Flex workers describe an identical experience. The algorithm decides. There is no human to argue with. This is enshittification with the friction of human management removed.

In AI-managed platforms dynamic pricing works against workers by design. The algorithm sets rates based on available labor supply. As traditional employment contracts and more people turn to gig work, labor supply grows and algorithmic rates fall. More workers competing means lower rates for all of them.

The platform extracts the difference. This is the Enshittification of the labor market itself. Workers were attracted by good rates in stage one, became dependent on the income in stage two, and are now trapped in stage three competing against each other for a shrinking pool of algorithmic compensation.

Why It Keeps Working on Us

The most effective feature of Enshittification is how gradual it is. No single change is bad enough to make you leave. The price goes up $2. The interface gets one more ad. The algorithm surfaces slightly worse results. The delivery window expands by 30 minutes. None of these individually cross the threshold that makes you cancel. But collectively, over years, the product you're paying more for has become genuinely worse than the product you originally signed up for. You stayed because of switching costs, habit, network effects, and the knowledge that the alternative is probably going through the same process.

The generation that grew up with these platforms is also the generation that watched every institution it trusted degrade in similar fashion. The cable company that was the only option. The university that kept raising tuition. The healthcare system that kept raising premiums. Enshittification isn't a tech sector problem.

It's what happens when monopoly or near-monopoly power meets a business model that prioritizes extraction over value creation. We just finally have a word for it.

Stay Frustrated