Against the feature matrix
Every buyer's guide rewards the longest checklist, so every product grows one. We think that race is exactly backwards, and it's why most software feels worse the more it can do.
There is a spreadsheet somewhere, there is always a spreadsheet, with your product's name in one column and a competitor's in the next, and a long list of features running down the side. Green ticks. Red crosses. Whoever wins the most rows wins the deal. So every team learns the same lesson early: add rows, never be the column with the gap. We've decided not to play that game, and this is why.
The matrix measures the wrong thing
A feature matrix records presence, not quality. It asks can it? and never once asks is it any good? On paper, a tool with forty adequate features beats a tool with four excellent ones every time. But nobody lives on paper. On the actual desk, in the hands of someone with a real job to finish before lunch, the forty-feature tool is the one that buries the thing they need under thirty-six things they don't.
The matrix can't see that. It has no column for "gets out of the way", no row for "I understood it in a minute", no score for the quiet confidence of a tool that does exactly what it says. Those are the only things that matter once the deal is done and the work begins, and they're precisely the things a checklist is blind to.
Every feature is a tax
The cost of a feature is not the week it takes to build. That part is cheap and everyone budgets for it. The real cost is the tax it levies forever after: another item in every menu, another paragraph in every doc, another concept in every onboarding, another edge case in every future change. Each feature makes the next one harder to design, because now it has to fit around what's already there.
"Just in case" is one of the most expensive phrases in software. A feature added for a handful of buyers is paid for, every day, by everyone else: in attention, in clutter, in the extra half-second it takes to find the one control they actually came for.
The question is never "would someone use this?" Someone will use anything. The question is "is it worth what it costs everyone else?"
Saying no is the product
Design is subtraction. The tools people genuinely love are defined as much by what they refuse to do as by what they do; the confidence to leave things out is the whole discipline, not a failure of nerve. We say no a great deal. No to the feature that demos beautifully and rots quietly. No to the setting that exists only so we never have to make a decision. No to the integration that adds a logo to the website and a liability to the codebase.
None of that is easy, and none of it shows up on the matrix. Restraint is invisible on a checklist and obvious in the hand. We'd rather lose the spreadsheet and win the Monday morning.
Before anything ships, it has to pass one plain question.
Does this respect the person on the other side of the screen: their time, their attention, their ownership of their own work? A feature that fails that test doesn't ship, however well it demos. That's what "always kind" actually means: not a tone of voice, a gate everything has to pass through.
What we build instead
Small machines. Each one does a single job better than anything else and then stops. Braidsheet is a spreadsheet you can actually read, a model whose logic is legible instead of hidden behind a thousand opaque cells. VulkanMM connects two second-brain vaults directly, peer to peer, with no server sitting in the middle owning your thinking. Neither tries to be a platform. Both try to be the best possible version of one thing.
That's the shape of the studio: a family of sharp, small tools, not one blunt platform that does everything adequately. When the two overlap, that's where the taste lives. Where they don't, we're happy to point you somewhere else.
The one you can always leave
There's a last column the matrix never includes, and it's the one we care about most. We'll charge for some of these tools, and some will run as a service, and that's how a studio keeps the lights on and keeps building. What we will never do is hold your work hostage to it. Your data stays yours: export anytime, in a format you can actually use, with no lock-in and no exit fee. The door is always unlocked.
The feature matrix will always have a row we lose, and we've made our peace with that. We're not building the product that wins the spreadsheet. We're building the one you're glad to open on a Monday, and the one you could walk away from tomorrow, if we ever stopped earning the place. That's the deal. That's kind.