The removal economy

In the early 2000s, airlines discovered something remarkable. You could take a product people already paid for, break it into pieces, and sell each piece back to them separately.

Checked bags. Seat selection. Meals. Legroom. Boarding priority. Overhead bin space.

The flight itself got cheaper. Everything around it got more expensive. And here's the part that made the whole thing work: the core product didn't change at all. A plane still took you from one city to another. You still sat in a chair. You still arrived. But somewhere between 2001 and 2010, "a flight" stopped being a product and became a platform for selling you back things you used to get for free.

The real genius was the psychology of it. Passengers used to complain about the in-flight sandwich. It was terrible. Everyone agreed. But when they started charging $9 for it, people bought it anyway. Turns out a bad sandwich you got for free was an annoyance. A bad sandwich you chose to buy was a purchase. Same sandwich. Different revenue line.

Tech companies watched this happen. And they learned. But they got better at it, because when an airline removes your free checked bag, you notice immediately. It's right there on the receipt. When a phone manufacturer removes your SD card slot, they call it courage. When a SaaS company kills a free tier, they call it simplification. When a laptop maker designs a battery you can't replace, they call it innovation.

The most profitable product decisions of the last decade haven't been about building new things. They've been about removing old things and selling the replacement. There's a name for this pattern worth remembering. Call it the Removal Economy. And if you build products for a living, you've almost certainly been in the room where it happens, even if nobody called it that.

Removing a feature is more profitable than building one

The economics of removal are strangely elegant once you see them clearly.

When you build a new feature, you add cost on every dimension. Development. QA. Documentation. Support tickets. Maintenance. Cognitive load in the product. Server costs. Every new feature is a small ongoing tax on the entire organisation.

When you remove a feature, you subtract all of that cost. But here's the part that makes CFOs smile: if the removed feature created a dependency, you can now sell the replacement. You're improving margins on both sides of the ledger simultaneously. Costs go down. Revenue goes up. No new feature in history has that return profile.

Consider this. Apple removed the headphone jack from the iPhone in 2016. AirPods launched the same year. By 2024, the wearables and accessories category (driven largely by AirPods) was generating over $23 billion annually. The removal of a component that cost less than a dollar to include created a product line worth more than most companies. That's not a side effect. That's the strategy.

I watched this playbook run at a SaaS company I worked at a few years ago. The product had a free reporting dashboard that was, frankly, pretty decent. Users liked it. Support tickets were low. It worked. But the margins weren't there, and the analytics team had built something better behind a paywall.

So the free tier disappeared in Q2. The paid analytics add-on launched in Q3. Same roadmap. Same planning cycle. The internal pitch never once used the word "monetise." It was all about "simplifying the core experience" and "reducing surface area." Revenue went up 18% that quarter. In the all-hands meeting, they called it a product improvement. In the finance review, they called it what it was.

The most honest sentence in product strategy is: "We removed it so we could sell it back to you." But nobody ever says that in the meeting. They say "simplify."

The difference between courage and extraction is about $159

Not all removal is the same. And this is where the conversation usually gets lazy. The viral post that kicked off this discussion (43,000 likes and counting) listed a string of "peak capitalist" product decisions: SD card slots, headphone jacks, chargers, repairability. And the implication was clear: companies are greedy, users are getting squeezed, tech is broken.

But that framing misses something important. Some removal genuinely makes the product better. The trick is knowing which is which.

There's a spectrum here, and it has two ends.

On one end: legitimate simplification. The removed feature is replaced by something better, at no additional cost to the user. Apple killed the floppy drive in 1998. That was foresight. The floppy was dying, USB and the internet were already better, and users gained more than they lost. Nobody mourns the floppy drive. Nobody mourns the physical keyboard on smartphones, either. The touchscreen was a genuine leap. These removals made the product better, full stop.

On the other end: extraction. The removed feature is replaced by a purchase. The user's experience stays the same or gets worse, unless they open their wallet. The headphone jack disappears, replaced by AirPods at $159. Chargers vanish from the box, sold separately for $19. SD card slots are eliminated, replaced by iCloud storage at $2.99 a month, forever. Repairability is engineered out, replaced by AppleCare subscriptions.

The diagnostic is simple. When a company removes something and the replacement is free, that's progress. When the replacement has a price tag, that's the Removal Economy.

It's the equivalent of a restaurant removing the bread basket from the table and adding a $4 "artisan bread selection" to the menu. The bread didn't get better. The margin did. And the waiter will tell you it's about "elevating the dining experience."

Every single example in that viral post falls on the extraction side of the spectrum. That's not a coincidence. That's a pattern. And the fact that 43,000 people recognised it instantly tells you something important about where trust stands between tech companies and the people who use their products.

Courage is removing something your users don't need anymore. Extraction is removing something your users still need and charging them to get it back.

The meeting where nobody says "should we?" because everyone is asking "can we?"

Here's where this gets uncomfortable for anyone who builds products for a living. And that includes me.

Removal decisions don't feel predatory from inside the room. They feel rational. There's a business case. There's a slide deck with projections. There's a metric that goes up and to the right. The PM presenting the proposal isn't twirling a villain's moustache. They're doing their job. The problem isn't that someone in the room is evil.

The problem is that the room is optimised to answer two questions: "Can we do this?" and "What's the revenue impact?" Nobody's job depends on asking "Should we do this?" and "What's the trust impact?" Those questions don't have a row in the spreadsheet. They don't show up in the OKR review. So they don't get asked.

I've been in one of these rooms. A product I worked on had a free API integration that a chunk of our user base relied on. It wasn't glamorous. It wasn't growing. But it worked, and people depended on it. The team proposed sunsetting it and routing users to a paid marketplace instead. The business case was clean. The integration was expensive to maintain. The marketplace had better margins. The PM presented it as "retiring a legacy feature."

Three people in the room had reservations. I was one of them. Nobody raised them, because the VP had already endorsed the direction and the quarterly targets were aggressive. You learn to read rooms in this industry, and this room had already decided.

Short-term revenue spiked. The next two quarters looked great. But twelve months later, two of the largest enterprise customers had quietly built their own workaround and were actively evaluating competitors. They didn't leave because the product got worse in some measurable way. They left because they felt the product was no longer on their side. One of them told our account manager, almost exactly: "We stopped trusting the roadmap."

The removal funded one quarter. The trust loss cost three years of pipeline.

In my experience across multiple companies, the decision to remove is almost never framed as a trade-off. It's framed as an optimisation. And optimisations don't require the same scrutiny as trade-offs. They get approved, not debated. That framing gap is where the real damage happens.

The most dangerous product meeting is the one where everyone agrees it's just an optimisation. Optimisations don't get debated. They get approved.

Every removal is a withdrawal from the trust account, and trust accounts don't send low-balance warnings

Every product relationship runs on a trust account. Features, reliability, fair pricing, the feeling that the company is building for you and not just from you: these are deposits. Every removal that benefits the company more than the user is a withdrawal.

But trust accounts don't work like bank accounts. There's no balance notification. No overdraft alert. No friendly email saying "Your trust balance is running low, please make a deposit." Users absorb one withdrawal. Then another. Then another. They don't complain proportionally. They don't send you a linear signal that things are getting worse.

And then they leave. Or they post something that gets 43,000 likes.

It's like how bridges fail. They don't sag gradually and give engineers time to reroute traffic. They hold, and hold, and hold. And then they don't. Structural engineers call this brittle failure. Product teams experience it as churn that "came out of nowhere." But it didn't come out of nowhere. It came out of every small withdrawal nobody was tracking.

If you're a PM, a designer, or a founder reading this, the Removal Economy isn't just a Big Tech problem you can point at from a safe distance. It's in your product, too. Every subscription tier you "simplify," every free feature you sunset, every integration you deprecate, every pricing change you frame as an improvement: each one is a decision on this spectrum.

The question isn't whether the revenue math works. The revenue math always works in the short term. Removal is the easiest win in product strategy. That's precisely why it's dangerous.

The question is what you're withdrawing from the trust account. And whether you can afford the balance when it comes due.

Trust accounts don't send low-balance warnings. They just close.

The airlines unbundled everything. Bags, seats, food, legroom, boarding order, overhead bin space, the right to bring a coat. They won every spreadsheet. They optimised every revenue line.

And now flying is something three billion people endure, not enjoy. The product is cheaper on paper. The experience is hollowed out. The margins are better than ever.

That's the Removal Economy's final act. You optimise the revenue model until there's nothing left worth paying for.

The question for anyone building a product today isn't what else you can remove. It's what you can't afford to.