Things 3 is a finished object in a business of services

In the spring of 2018 I stood at the back of an Apple Store, waiting on a repair that was never going to happen, and listened to a couple conduct a procurement review over a display unit. The woman had a list in her hand, the kind typed into Notes, and she asked the blue shirt the only question that mattered: could she put the shopping list in the app and have him see it on his phone. He said the app was really designed for one person, and the pause after that was the length of a marriage doing arithmetic. She asked what it cost. Fifty dollars on the Mac, ten on the phone, he said, once. They bought two copies on the spot, of an application that could not do the one thing they had walked in for, and neither of them appeared to think a strange transaction had occurred.

The scene was the argument. Things 3, on sale since May 2017, is the last major task manager that behaves like an object you own rather than a service you rent, and the things it refuses to do are not oversights. They are the balance sheet talking. Cultured Code is a small studio in Stuttgart that has spent nine years declining to grow its product, and the refusals are so consistent, so patterned, that refusing is clearly the strategy. Whether that is integrity or abdication is the actual question, and I intend to weight it properly.

A one-time price is an incentive, and it points at the purchase

Follow the money, because the money explains the software. A subscription company's revenue arrives monthly and is earned by staying useful, or at least by staying open; a purchase company's revenue arrives once, at the moment of decision, and is earned by being lovely at that moment and by nothing else. Things 3 is what that incentive produces. The design was effectively complete at launch, polished enough to win an Apple Design Award the year it shipped, and the Mac, iPad, iPhone and Vision Pro prices ($49.99, $19.99, $9.99, $29.99) have sat where they landed in 2017. No price creep, no paywall between you and reminders, no AI panel asking to be noticed, and sync included free, which no subscription competitor can say without a footnote. PCMag's 2025 review, by Ruben Circelli, scored it 3.5 and docked it for the per-platform pricing and the missing platforms, which is fair grading of a product category and slightly beside the point of this one.

I bought my copy in 2017 and it is open on the Mac as I write, which makes me the incentive's favorite subject: someone who paid once and has been served quietly ever since. I want to be honest about what that feels like from inside. It feels like a tool that is not thinking about me at all, and I have come to understand that this is the luxury. Every other application in my dock is thinking about me constantly, in the specific sense of measuring me, and Things 3 is the one surface that has never once tried to convert my attention into revenue.

The missing features are the balance sheet

Now the inventory of refusals, because the list is the argument made in negative. No shared projects, no assignment, no comments, no collaboration of any kind. No public API for automation. No web app, no Windows client, no Android, nothing outside the Apple walled garden. Every one of those gaps is a roadmap item at a subscription company and a permanent liability here, because a feature is not a launch, it is a maintenance surface you fund across four platforms forever, and the money to fund it stops arriving the moment the customer stops buying. Cultured Code looked at that math and declined the whole column at once. The community has noticed, in the way communities do. There is a thread on the Things subreddit titled around when Cultured Code will start listening to its existing customers, another called, I promise, "Cultured Code's laziness is finally catching up with Things 3," and a recurring joke that GTA 6 will ship before Things 4. Both things are true. The pace is genuinely neglectful by the standards of 2026 software, and it is also the reason the app still opens in 2026 exactly as fast as it did in 2017, with nothing added between you and your list, and nothing taken away.

Consider the incentive, stated without euphemism. A company that cannot charge you again has no reason to make you unhappy, and also no reason to make you delighted. It has a reason not to break, and a reason to stay lovely at the margins where a new buyer might look, and nothing beyond that. Nine years in, the strategy has a certain cold perfection: the customers who wanted the service model left for it, the ones who wanted an object stayed, and the app itself has never once been the subject of a migration panic.

The relic argument, made properly

The case against deserves better than the sniping on Reddit, so here it is at full strength. Software is not furniture. A task manager lives inside an operating system that changes annually, above frameworks that deprecate, against servers it does not own, and a paid-once product with no revenue line has no budget for the day the environment moves underneath it. The Apple-only bet compounds the risk: eighty dollars across three devices, and the whole library is hostage to one vendor's decisions about APIs and services. And the collaboration gap is not a missing nicety. Task management for working adults is substantially the management of tasks other people hand you, and a tool that cannot accept a task from another human is a diary with a synchronizing spinner. The strongest version of the argument says the one-time model only works while the platform stays quiet, and that Cultured Code has not built a business so much as borrowed one from Apple's stability. App Store reviewers make the other complaint in a sentence: paying three times for one app "feels like a scam," and eighty dollars is not trivial against a $60 subscription that at least pretends to cover everything.

I find this argument close to conclusive about the future and beside the point about the present, which is a contradiction I am willing to hold. The object has survived nine years of platform weather so far. The diary objection is the one I would not want to argue against, because it is correct about most people, and the people it is not correct about are a niche, and niches do not fund a studio in Stuttgart. They do, apparently, fund one.

So the verdict, conditional as promised. If your work is solo, your machines are Apple's, and your failure mode is capture followed by action, Things 3 is the best-made object in its category and the price is a dinner, not a dependency. If you share work with anyone, or a Windows laptop follows you home from the office, its perfection is local to a room you are only sometimes in, and you should buy the subscription you need instead of the object you admire. The variable worth watching is not Things 4, which will arrive or won't. It is Apple, because the app has staked nine years of one-time revenue on a single company's interfaces staying still, and an object has no budget line for the day they don't. The blue shirt was right about the sharing. They bought two copies anyway, which is the whole review.