Three Moments When the Product Stopped Being Yours

Your printer has worked flawlessly for three years. You've been using third-party cartridges that cost one-quarter what HP originals do. One day, it simply stops. An error message appears - "damaged or incompatible cartridge" - but the prompt is silent about the fact that a firmware update has just intentionally blocked any cartridge without an HP chip. It takes an hour of digging through technical documentation to realize what actually happened. The hardware hasn't changed at all. The agreement you made when you bought it, however, no longer exists.

You're playing a video game you bought. Not a subscription - you paid the retail price and received the title outright. The Crew, Ubisoft's open-world racing game, existed for nearly a decade. Then Ubisoft shut down the servers the game couldn't function without, and revoked purchased licenses. No refunds for the vast majority of players. The announcement was terse: servers are closing, support ends on a specific date.

For years, you cast Netflix to your TV with a single tap - smooth, effortless, part of your daily routine. In late November and early December 2025, an app update quietly removed that feature from most platforms. The official support page offered no explanation. A support agent told one user it was done to "improve the user experience." Your experience got worse, but your opinion wasn't part of the equation.

These three products have almost nothing in common - different categories, price points, use contexts. But they share one thing: they each started as a device or a one-time purchase and ended up as a service whose terms the manufacturer changed unilaterally. Ownership became a subscription - retroactively, without your consent.

Every one of those decisions had an author. Supporting data. A deadline. And someone who signed off on it - then drove home that evening and ran into the same question every user runs into: why can't I just play a game I paid for? The mechanism behind it doesn't require cynicism. It only requires an environment that rewards the right metrics.

A Pattern Someone Has Already Measured

The examples above are no coincidence. In February 2026, the Norwegian Consumer Council (Forbrukerrådet) published a report titled Breaking Free: Pathways to a Fair Technological Future, documenting this phenomenon across multiple markets simultaneously - social platforms, streaming services, and internet-connected home devices. The report doesn't treat it as a series of missteps by individual companies. It frames it as the expression of a single system - one that has eliminated the natural brakes on honest behavior.

In theory, you can cancel a subscription or find a different printer. But the device you already paid for is sitting on your desk and won't print. And the longer you use a platform, the more it knows about you, and the harder it becomes to leave. In that structure, manufacturers can change the terms at will with no consequences - and honest behavior stops being profitable.

The Cycle Cory Doctorow Described and Named

The term enshittification was coined by writer and technology activist Cory Doctorow. It describes the cycle in which a platform or service first delivers genuine value to attract users - then, once those users are locked in, gradually degrades their experience in the pursuit of greater profit. Although Doctorow coined the term in the context of internet platforms, both he and Forbrukerrådet apply it more broadly, including to physical devices whose software becomes the instrument through which usage terms are quietly rewritten.

The cycle has three phases. In the first, the product works and delights - users flood in. In the second, once the user base is large enough, the company begins favoring business customers at users' expense. Search results give way to ads. The algorithm stops surfacing what you're looking for and starts surfacing what advertisers paid to promote. In the third, once both groups are sufficiently locked in, the company harvests value from everyone at once: users get a worse experience, business customers pay more for diminishing returns, and the difference flows to shareholders. Search engines, marketplaces, social media - all have moved through this cycle.

Doctorow's model describes two-sided platforms, with users on one side and business customers on the other. With devices and one-time purchases, the mechanism is analogous but simpler: there's no intermediary. There's a manufacturer, embedded software, and a user who can't walk away. The cycle plays out identically - just without the network effects that, on platforms, can accelerate user response. A firmware change is silent. You only notice it when the printer stops printing.

Doctorow himself points to specific exit routes: restoring interoperability between ecosystems and enforcing antitrust law. In other words - the fix has to come from changing market structure and regulation, not from individual consumer choices.

But where exactly does it begin?

The Sprint Review

Picture a typical product meeting. Every two weeks, the team - PM, designers, engineers, analyst - sits down and reviews what got shipped. Nothing unusual. These meetings happen in thousands of companies worldwide.

On the screen: A/B test results. This time, it concerned what happens when a user tries to cancel their subscription. The data is unambiguous: users who encountered an "are you sure you want to leave?" screen during the cancellation flow abandoned the process 23% of the time. The change produced a measurable lift in retention.

One of the designers says it plainly: this screen looks like it was designed to obstruct the decision, not support it. The "Stay" button - larger, brighter, up front. The "Confirm Cancellation" button - tucked at the bottom of the screen, grey, requiring an extra scroll. That's a dark pattern: a deliberate friction that steers users toward decisions that don't serve their interests. In the practice of product design, the line between optimization and manipulation is rarely obvious in the moment a decision is being made.

Nobody says, "let's do something unethical." The PM says, "the data shows this works." The engineer says, "this is standard for this product category." Someone mentions the release notes deadline is the day after tomorrow. The feature ships. The decision gets made - not because anyone wanted to harm the user, but because the decision-making environment rewarded the outcome, not the intent.

This is exactly where enshittification begins - not in the boardroom, but in a conference room, with clean-looking data and a deadline two days out.

The problem isn't the people. It's the environment they're working in. Pressure on quarterly results and optimization for short-term metrics - retention rate, conversion, session length - creates conditions where every small decision made at the user's expense is structurally rewarded, even when no one consciously chooses it. Taken individually, each decision sounds reasonable. Together, they build a pattern no one planned - because everyone only signed off on their piece. The Forbrukerrådet report puts it directly: "although enshittification is the result of deliberate choices made by individuals within companies, developers and other employees often have little power to push back against the drive to maximize profit."

The scale of the phenomenon is reflected in the data. An FTC report from July 2024, covering 642 subscription services across 26 countries, found that nearly 76% employed at least one dark pattern, and nearly 67% used multiple simultaneously. The most common: automatic renewal with no opt-out mechanism, and deliberately obstructed cancellation flows. These are not edge cases. They are the norm.

Data determines what stays. Data approves what ships. And data is collected from systems that measure short-term outcomes - with no field for the question: is this change fair to the user?

That distinction matters. And it's what makes this problem so resistant to individual solutions.

When the Victim Is a Designer

Figma is the interface design tool used by the overwhelming majority of professional UX designers worldwide. In 2024, the company automatically enrolled users on Starter and Professional plans into a program that allowed their design files to be used to train AI models. An opt-out was available - but buried in team admin settings, far from where an average user would expect to manage data consent.

The backlash was fierce, precisely because it didn't hit ordinary users - it hit designers. People who professionally understand what a hidden opt-out looks like, and what default consent means in an interface. In November 2025, a class action lawsuit was filed in federal court in California, alleging the unlawful use of customer data to build AI tools that directly contributed to the company's valuation ahead of its planned IPO. The decision to make content training opt-out by default - and its consequences - remains the subject of active litigation.

The pushback worked - up to a point. But that doesn't change the fact that this decision was made in the first place - by people who understood the mechanism from the inside.

If an environment that professionally understands these patterns is not immune to them, that's a signal that the fix has to come from structural change: regulation, certification, and institutional pressure that shifts the cost of bad design decisions from the user to the company. That change is coming.

The Reckoning

Something real is shifting. Not as a voluntary course correction by principled companies - but as a change in the market and legal conditions in which product degradation is no longer cost-free.

Stop Killing Games started as a campaign by players demanding that publishers not be allowed to permanently disable purchased games by shutting down servers. It grew into a European Citizens' Initiative with over 1.29 million verified signatures - a legal instrument that obliges the European Commission to formally respond to the issue raised. This is not a petition. It is formal institutional pressure, triggered directly by Ubisoft's shutdown of The Crew in 2024.

Amazon's September 2025 settlement with the FTC - $2.5 billion in total, comprising a $1 billion fine and $1.5 billion in refunds to 35 million affected consumers - addressed exactly what, in design terms, we call a problematic cancellation flow: a Prime subscription cancellation process designed to be confusing, multi-step, and discouraging. This is no longer a question of design aesthetics. It is a legal exposure measured in billions of dollars.

The HP case is particularly telling. For years, the company justified blocking third-party cartridges on security grounds. The argument held for an entire decade. Then, in December 2025, EPEAT 2.0 standards took effect - an environmental certification for electronic hardware on whose Gold certification HP had built its narrative of environmental responsibility for its computers and monitors. The new standard explicitly requires certified printers to accept cartridges remanufactured from original components. In January 2026, HP released firmware 2602A/B, extending the block to eleven additional printer models. The security argument backfired on the company - not because of a consumer campaign, but because of a certification standard whose logo HP had been using in its own marketing materials as proof of environmental stewardship.

More is coming. The Right to Repair Directive (EU 2024/1799), adopted by the European Parliament in June 2024, requires member states to implement it by July 31, 2026, and covers electronic devices with embedded software. For teams designing such hardware, firmware architecture has ceased to be a purely technical decision - it is now a compliance question. The Digital Fairness Act, currently under consultation at the European Commission, is set to directly regulate dark patterns in digital interfaces. The final text hasn't been adopted yet - but design decisions made today will land in products deployed under an entirely new legal environment. Waiting for the final text will mean expensive redesigns of finished products.

Those Who Chose Differently - and a Case Study in Failure

Fairphone positions repairability as a sales argument, not a regulatory concession - and is growing: revenues rose 35% in 2025. GOG has consistently held up DRM-free gaming as its core differentiator, even after a change in ownership in 2025. Framework has built its market position in a segment that spent years treating service as a margin opportunity rather than customer value - and demonstrates that decisions made as early as the prototyping and product development stage determine whether repair will even be possible at all.

None of these examples is infallible, and none guarantees success. But each answers the same strategic question: is treating users fairly a cost, or a positioning strategy?

Sonos is the inverse of that question.

In May 2024, Sonos - a maker of wireless speakers with a loyal customer base and a firmly established market position - launched a redesigned mobile app alongside its new Sonos Ace headphones. The new interface was visually cleaner. It also removed a range of features that users relied on daily: sleep timers, alarms, and key accessibility options for users with disabilities. Communication about what was gone and when it might return was inadequate. What followed were bug reports, delays, and accounts of older Sonos devices that stopped working altogether.

Fiscal Q4 2024 results told the story: a 16% quarter-on-quarter revenue decline. The company's stock lost 13% of its value following the update's rollout. In August 2024, a hundred employees were laid off. In January 2025, CEO Patrick Spence resigned.

Sonos users didn't discover that the company had been acting in bad faith. They discovered that it had stopped listening to them. That was enough.

Three Questions Before Approving Any Change

These are tactical tools, not a substitute for structural reform. But they can delay the moment when fixing a product becomes prohibitively expensive - or simply impossible.

First: is this a technical constraint, or a business decision dressed up in code? The test is simple: can this change be described honestly in an update notice directed at the user? If the message to users would have to say something different from what the change actually does, that's not a communication problem. That's a design problem.

Second: will the promise made to the user hold up to outside interpretation? Could someone looking only at the product page describe what they're getting - and what might change after purchase? If an honest answer requires reading the terms and conditions in the footer, that's not an answer.

Third: is this change better for the user, or does it just move the metrics? Ask someone outside the project to describe, in one sentence, what this change means for the user. If the answer describes the company's outcome rather than the user's experience - it's not the right answer yet.

Before Someone Approves the Next Change

Enshittification doesn't start with cynical calculation. It starts in environments that fail to ask the right questions before a pattern becomes irreversible.

One sentence at every change is enough: is this better for the user, or does it just improve our numbers? And let that answer make it into the release notes.