Four Years of Writing This Journal

We published the first post in this journal in the summer of 2022. The plan was modest: write down what we actually believed, roughly once a week, and let the archive become a record of how a small studio thinks. Four years later, the archive is the closest thing we have to an institutional memory, and I want to mark the anniversary by being honest about what the habit did to us.

Writing forced us to have positions

The first thing you learn when you write in public is that vague opinions do not survive contact with a blank page. "We care about craft" is a poster. "We will not ship a homepage that takes four seconds to load, even if the client stops noticing after the second review" is a position. The journal made us convert our taste into commitments, and those commitments quietly became our operating system.

Some of them aged well. We argued in 2022 that small senior teams would beat big shops on both quality and speed, back when that was a contrarian pitch rather than the default assumption of every procurement conversation. We argued that measurement theater was rotting media budgets. We argued that the studios who survived the AI wave would be the ones who treated it as a material, not a miracle.

Some of them did not age well, and we have left those posts up. In early 2023 I wrote that generative tools would stay in the moodboard phase for years. That prediction lasted about eight months. Leaving the miss visible matters more to me than looking clairvoyant. A journal that only contains correct predictions is a portfolio of edits.

What changed between 2022 and now

Reading the archive front to back over the holidays, three shifts stand out.

• In 2022 we wrote about tools. By 2024 we were writing about workflows. Now we mostly write about judgment, because the tools and workflows stopped being the constraint.
• Client questions inverted. Four years ago the anxious question was "should we be doing this AI thing?" Today it is "everything we make looks like everyone else's AI thing, can you fix that?"
• Our own posts got shorter and more opinionated. Early entries hedged. We do not hedge much anymore, partly because the archive gives us receipts.

The through line is that taste became the scarce resource. Everyone can produce now. Very few can decide. The journal, it turns out, was four years of practicing deciding in public.

The rules we kept

People occasionally ask how we sustain a weekly-ish cadence without it becoming content marketing sludge. The answer is a short list of rules we have never written down until now.

Every post must come from real work or a real argument we had in the studio that week. No post exists to rank for a keyword. Anyone on the team can write, but nobody is forced to, and we would rather skip a week than publish filler. And every post has to risk something: a prediction, a criticism, a price, a number.

That last rule is the whole game. A journal with no risk in it is a brochure.

What the next year looks like

2026 is already shaping up to be the strangest, busiest year in the studio's history. A Super Bowl in our own backyard next month. A Winter Olympics the same week. A World Cup arriving on this continent in June, with half our clients planning around it. Agent-driven commerce moving from conference talk to roadmap item. And a pipeline that, for the first time ever, has us having conversations about Q4 before we have finished putting away the holiday decorations.

We will write about all of it here, the way we always have: from inside the work, with our actual opinions attached. Thank you for reading for four years. The archive is open, misses and all.