A while back we took on a client whose product we didn't really believe in — the timeline was tight, the budget was good, and we told ourselves we'd figure out the "believing in it" part along the way. We didn't. The project shipped, it worked, and it was some of the least proud we've ever been of our own output. That's the moment this company actually started taking its own manifesto seriously.
Saying yes was the easy habit to build.
Early on, every "yes" felt like survival — revenue, a reference client, proof we could deliver. What we didn't account for was the cost of a "yes" said to the wrong project: the good clients get less of you, the team gets stretched across things nobody's excited about, and the work quietly gets worse across the board. Saying no turned out to be the harder, more valuable skill, and it took longer to learn than we'd like to admit.
The AI hype cycle taught us more by watching it than by chasing it.
For a stretch, it felt like every client conversation started with "we need AI in this" before anyone had asked what problem it was solving. We built a few of those features. Most of them got quietly removed within a year — not because AI doesn't work, but because it was solving a marketing problem, not a user problem. The AI work we're actually proud of is boring by comparison: it's grounded, it's evaluated, and nobody has to write a press release about it for it to be worth the investment.
Small isn't a phase, it's a decision.
For a while we assumed staying small was a temporary state on the way to something bigger. It's not — it's the actual shape of the business we want to run. Every hire beyond what the work needs adds distance between the person who talked to the client and the person who wrote the code. We've watched that distance be exactly where quality goes to die at other shops. We'd rather stay the size where that distance can't exist.
None of this makes us special. It just makes us consistent.
We don't think "build good software, tell clients the truth, and don't take work you can't be proud of" is a novel insight. It's closer to the obvious thing everyone agrees with in principle and very few teams actually hold themselves to under deadline pressure. That's the whole bet this company is making — that consistency on the obvious stuff is rarer, and more valuable, than it should be.
Want the honest version of a project's chances?
That's the only version we know how to give — on this project or the next one.