A method, not a manifesto.
None of what follows is unusual on its own. Holding all of it at once, on everything that carries our name, is the part that takes discipline — and most of the reason the work comes out the way it does.
- What we take on01
Difficulty is the filter
We choose by how hard something is to do properly. A problem anyone could solve does not need us, and one that cannot be solved well is not worth the years it would take. What remains is a short list, which is the point.
- How it should feel02
Advanced technology, quietly applied
We work at the front of what machine learning can currently do, and we consider it a failure if anyone using the result has to think about that. The measure is not that the technology is impressive. It is that the product behaves correctly and the question never comes up.
- Where quality lives03
Detail is not decoration
Typography, latency, the phrasing of a single sentence, the moment a screen comes to rest: these are not finishing touches. To the person using the product they are the product, and they are decided long before anything is finished.
- What we leave out04
Restraint
Most ideas are refused, including good ones. A smaller amount of software held to a higher standard outlives a larger amount held to none, and every feature that is not there cannot age badly, break quietly, or dilute the rest.
- How we treat data05
Confidentiality by construction
What our software learns about a person belongs to that person. We design so that the shortest, least exposed path is the one the product takes, and we hold the least we can rather than the most we are permitted.
- The time horizon06
Built to last
We write expecting the software to be running years from now, and we keep supporting what we publish. That expectation changes what gets written, what gets thrown away, and what is accepted as good enough on any given afternoon.
Questions are welcome.
If you want to know how a particular piece of this works, ask. We would rather explain it than have it taken on faith.