One Project at a Time: The Attention Economy of Creative Work

Three projects running at once, each at thirty-three percent, look equivalent to one project at a hundred. They are not. That arithmetic works for moving bricks, not for work where the result depends on how deep into the problem you managed to get.

Technical and creative work carry a fixed entry cost. To write a good function inside a system, you have to hold the system in your head: how the data moves, which conventions were settled, which compromises were made three weeks ago and why. Rebuilding that takes time. On a project of moderate complexity, somewhere between twenty minutes and an hour — not to open the files, but to become again the person who understands why they look that way.

Switch context three times a day and you pay that tax three times. What is left of eight hours is not eight hours of work but five of work and three of warming back up. And the warm-up hours are not neutral. They are where the bad decisions get made, the ones taken on half a context and under pressure to show somebody progress.

Context switching does not only steal time. It steals depth. Time can be made up with longer days; depth cannot.

There is something else that never shows up in an estimate: the good ideas do not arrive during the work, they arrive after it. They come on the third consecutive day spent inside the same problem, once the mind has had a chance to run it in the background and hands you the answer while you are doing something else. A fragmented schedule never reaches day three.

This is not a moral criticism. It is a description of an economic model. An agency with fifteen active projects needs a mechanism that makes fifteen projects possible, and that mechanism has three parts.

• A layered team. The senior you met in the pitch does architecture and meetings. Execution goes to people who cost less per hour. It is financially rational, and it is usually invisible to the client until the first strange behaviour in production.
• Templates. A reusable component set lowers the marginal cost of every new project. The result is a site that works but that resembles, structurally, the other twenty shipped this year.
• The queue. Your project is not a priority; it is a position in a list. When someone else has an emergency, you are the adjustable variable. That is why the answer to a simple question arrives four days later.

What the client loses is usually not raw code quality. It is the things that cannot be delegated: coherence of decisions over time, memory of the project, the instinct that today's proposal contradicts something settled a month ago.

Zaiko works with one client at a time. That is not a marketing position — it is the only way to avoid the mechanisms above.

Speed of decisions

In a structure with an account manager, a question makes four trips: client, account, developer, back. Every trip adds latency and loses nuance. When you talk directly to the person writing the code, an architectural decision gets made in ten minutes, inside the same conversation where the question came up.

One accountable mind

There is no context handoff between designer and developer, because there is no handoff. Design decisions know the technical constraints at the moment they are taken, and technical decisions know what the design is after. The discussion about what a website actually costs deals with the numbers; this one is about what happens to them once they are approved.

Coherence of the result

A site built by one person has something teams struggle to reach: a point of view. The spacing, the tone of the copy, the scroll behaviour, the way a form reacts when you get something wrong — all of it comes from the same head. It does not read as style. It reads as the absence of friction.

There is no point presenting this as a choice without costs. It has three, and they are real.

• You wait for a slot. If the slot is taken, we cannot start tomorrow. You can see what is being built right now on the current project page, and join the list with an estimated start date rather than a vague promise.
• The cost per project is higher. There are no economies of scale and no junior hours pulling the average down. You are paying for undivided attention, and undivided attention is expensive. That is the whole explanation.
• You depend on a single person. This is the serious risk and it deserves to be said plainly. It is mitigated with documentation written along the way rather than at the end, code handed over in standard shapes with no proprietary inventions, full infrastructure access in your own name, and a handover guide that lets any competent developer pick it up.

Anyone who needs ten landing pages a month does not need me. That is a choice of instrument, not a verdict on quality.

When there is only one project, an entire category of conversation disappears: the one about availability. You stop negotiating attention, and you stop asking for estimates as a way of finding out whether you are a priority. The discussion moves to substance, and that changes the rhythm too: long working sessions with decisions made on the spot, instead of a steady stream of status reports.

The Japanese have a phrase, ichigo ichie — one meeting, one opportunity. The idea is not sentimental but practical: each encounter is treated as though it will not repeat, so it deserves your full presence. Applied to work, it means the project in front of you is not one among several. It is the only one.

I am aware of how this sounds. The waitlist is a classic sales instrument, often used with nothing behind it.

The difference here is checkable: the number of slots is not set by marketing, it is set by the arithmetic at the top of this article. If I took on three clients at once, I would be forced to adopt exactly the mechanisms described above — juniors on execution, templates, a priority queue. Scarcity is not the cause of the method. It is its consequence, and it disappears the moment the method changes.

In a market where supply is abundant and differentiation is weak, and where new tools keep lowering the cost of producing mediocrity, the only thing that stays scarce is judgement applied consistently to one thing. If you have a project that deserves that, write to me — I will tell you honestly when the next slot opens and whether we are a good fit.