The Switch is a category, not a console.

The Switch launched on March 3rd, three weeks ago. Nobody we know has stopped talking about it. Every review opens with the same paragraph trying to explain what kind of device it actually is — handheld? console? tablet? hybrid? — and every one of them gives up halfway through and just describes what the Switch does. Nintendo, by all appearances, are completely fine with that ambiguity. They have shipped something that does not fit an existing box, and the market is making room for it in real time.

We have a Switch in the studio. We've been passing it around. We're going to write proper product notes about a games console because the way it was built is more relevant to anyone shipping a product than most of the industry pieces we read in a month.

The Wii U story is the necessary prologue

The Switch did not appear out of nowhere. Nintendo spent four years selling the Wii U — a console with a touchscreen tablet controller and almost no commercial success — and the lessons from that failure are visible in every Switch design decision. The Wii U's tablet was tethered to the TV. You could play it without the TV being on, but only within Bluetooth range. The split was awkward; the messaging was worse. Nobody knew what the Wii U was. Most people thought it was a Wii accessory.

The Switch took the same fundamental idea — a screen you can pick up and a screen you can plug in — and committed to it. No tether. No range limit. The same hardware does both. The marketing pretends there is no question to ask. That confidence is the entire product.

What it teaches

• 01If the existing categories don't fit your product, the answer isn't to compromise it into one. It's to ship the product anyway and let the category form around it. Reviewers will struggle for a month and then settle.
• 02Hybrid is only confusing when the modes feel like trade-offs. The Switch makes every mode feel like the best mode for the moment — sofa, train, plane, friend's living room — and never makes you feel like you've lost something by switching.
• 03Hardware/software co-design wins. Joy-Cons that physically separate are a software story as much as a hardware one — the OS, the games, the launcher all assume it. You see this in Zelda's pause menu, in 1-2-Switch's whole premise, in the dock animation.
• 04Restraint about specifications. The Switch is a fraction of the GPU power of a PS4 Pro. Nintendo neither hides nor argues about it. They have refused to compete on the axis where they would lose.
• 05Constraints applied honestly become identity. The handheld mode forced screen-readable HUDs and short play sessions; that constraint shows up positively in every first-party game.

The Joy-Cons are the most underrated product decision of the year

The two halves of the Switch's primary controller are detachable, can each be used as standalone tiny gamepads, work in either hand, have IR and motion sensors, and slide into a grip that turns them into a more conventional pad. This is a product so over-engineered that it should not work. It does work, because every game on the system was either built assuming Joy-Cons exist or was given a clear non-Joy-Con alternative.

The lesson is not 'design clever hardware'. The lesson is: if you ship clever hardware without commitment from the software side, you have built a Wii U. The Switch's success is software co-investment, not industrial design.

The launch lineup question, and what it answered

Console launches usually involve a slow-build software lineup over the first year. The Switch launched with The Legend of Zelda: Breath of the Wild — arguably the best game Nintendo have made in fifteen years — and not much else. The 'thin lineup' criticism was loud in the run-up. It evaporated within a week. One generation-defining game on day one outweighs ten merely-good ones over twelve months.

Builders, take note. When you ship a new product shape, ship it with one extraordinary thing that uses the shape. Do not ship a wide-but-shallow set of demonstrations. The Switch is selling out everywhere because one game is forcing the conversation.

What it means for the rest of us

Most of the products we build for clients sit inside well-defined categories — 'CMS', 'booking system', 'dashboard'. The brief usually arrives pre-shaped. The vast majority of our work is doing the well-known shape well. That is the job; we like the job.

But once a year, maybe twice, a brief shows up that is secretly a category-creation brief in disguise. The client doesn't know it. The conventional shape will work, badly. The reflex is to push it into the nearest box. The Switch is a reminder that the right answer is sometimes to refuse the box, build the thing the brief actually wants, and trust that the language to describe it will catch up.

When you take that route, you need three things: total commitment to the shape, one genuinely great example of why the shape matters, and the discipline to not apologise for it during the launch window. Nintendo got all three. Most products that try and fail to create a category get one or two.

"

If you ship clever hardware without commitment from the software side, you have built a Wii U. The Switch's success is software co-investment, not industrial design.

The cynic's counter-argument

The Switch could still flop. Two-month sales mean very little. The Wii U launched well and unwound across three years. Nintendo's history is full of brilliant first acts that didn't survive contact with their second. This is a reasonable take and we'll know in eighteen months whether the launch enthusiasm sustains.

Our bet, on day twenty-three: it sustains. The product has too many quiet, hard-won design decisions behind it. The Joy-Con architecture, the dock as a passive HDMI bridge, the launcher tuned for short sessions, the game lineup road-mapped explicitly for the hybrid model — none of that is accidental. Nintendo are playing a game they have been losing for ten years, and they have come back to it with the kind of clarity that is rare in any large company.

Also: it's a delightful piece of hardware. Pick one up. Try the kickstand. The kickstand is terrible. That's also part of why we like it — every detail isn't perfect, and the bits that are perfect are perfect for a reason.