The arithmetic of hours.
A serious site is six hundred to nine hundred hours of work. Design, development, content structure, review, testing, release, support. This isn't an "agency estimate," it's an honest number after a couple dozen finished projects.
A three-person studio, working forty hours a week and taking four weeks of vacation a year, is about fifty-five hundred person-hours a year. Divide. You get four to six projects.
In practice — three or four. There have to be gaps between projects: concepts, essays, updating the design system, new tools, retraining on new stack versions. Without that, in a year and a half the studio gets slower, not faster.
No magic. Just arithmetic and honest planning.
What "three projects a year" means.
In practice it looks like this. One or two projects run in parallel, offset in phase — while one is in design, another is in development. A third is in preparation: brief, concept, contract. A fourth is in the queue, with a confirmed start date.
The calendar is booked a quarter out, sometimes two. When an email arrives — "we urgently need a site by August" — and it's April, we answer honestly: the nearest slot is February of next year. Half the clients leave. Half agree to wait.
It means we know all our projects for the next eight to ten months. Their names, their briefs, their deadlines. No "we'll figure out who's next." That's a calm that's hard to explain to people working in "grab it while it's offered" mode.
What the limit gives.
The main thing is attention. When we have three projects, none of them is compressed to "can we go faster." Decisions are made slowly, because there's time to make them.
What that means in practice. A designer doesn't choose an animation curve "because there's thirty minutes, need to decide." They choose it because they spent three days looking at it in the browser, on a mid-range Android, in real typography, in the live context of the other components. And decided this curve is the right one.
A developer doesn't choose a CMS "because it worked last project." They look at the client's specifics, the editors' scenarios, the future three years out. And in eight cases out of ten they choose Payload — but in two, something else. Those two cases are what make the difference.
Attention doesn't scale the way revenue does. Doubling revenue is a commercial task with many tools for it. Doubling attention is impossible. Attention is a finite resource of three heads in a studio.
You can't do it well and in parallel. You can do one of the two.
What it takes away.
The main cost of the limit is missed projects. Last year about sixty clients reached out. We took four. Declined thirty-eight. Passed fourteen to colleagues. Failed to reply in time to four (our fault).
Among the thirty-eight declines are several projects we'd have liked to do. A good brand, a clear task, a reasonable budget. But we're already booked, and saying "let's talk in a year" isn't honest.
That's the price the studio pays for its choice of format. In exchange — no "burning projects," no "escalations," no burnout two years in. Most studios burn out precisely because they physically can't say no. For us, saying no is a built-in function.
The queue as a product.
On our "About" page we write: "Booking from Q3 2026, 3 of 4 slots open." This isn't a marketing pose. It's the calendar.
When a client writes in April and hears "we'll start in February," some are scared — "so it's not relevant." Some are reassured — "so it's thorough." The second group is ours.
The queue, which we keep transparent, is a signal. That the studio isn't trying to "fit everyone in." That we won't take a project until we finish the current one. That our "yes" means "yes." In a world where most studios promise "we'll start in two weeks" and start in two months, that signal is a rarity the client reads as a premium trait.
It isn't a marketing position. It's a side effect of the arithmetic.
The price of the limit.
The three-projects-a-year format doesn't work if the average fee is below a certain threshold. For us that threshold is about two hundred fifty thousand rubles for the smallest format (premium showcase) and about one million for the mid format (a flagship site). Below that, we can't cover salaries.
It means we're inaccessible to half of Russian small business. And that's fine. To small business we can only offer a site builder or a good freelancer, and we recommend that without hang-ups.
In a world where "quality should be accessible to everyone," this position looks elitist. In fact it's simply honest: we don't have a cheap version of our product because we don't know how to make one without losing quality. A site builder isn't "cheap NOVA," it's a different product.
What doesn't scale.
Sometimes we're told: "take six projects, hire two more people." That's mathematically possible — but it means a different studio.
At five people you already need a project manager. At seven — a dedicated account. At ten — a dedicated HR. By fifteen you get an open-plan office, a Monday all-hands Zoom, a careers page with an energetic team photo. That's a normal agency architecture. But it isn't NOVA.
We studied this trajectory in our peers. In the second year of growth, decisions usually get harder to make; in the third — focus is lost; in the fourth — half the projects are done without the founders. By the fifth year most founder-designers are in a permanent correspondence about hiring, not about design.
We chose a different trajectory — deliberately. And every year we revisit the choice. So far, it holds.
Who it's for.
Not every client needs a site from a studio like this. If your marketing plan says "a landing page for a campaign, for two months, doesn't matter how" — we're not your format. Take Tilda and a good freelancer.
If your site is a multi-year investment in the brand, your only scalable medium, the face of the company for clients you've only seen on Zoom — then three projects a year starts to sound different. It means that when your turn comes, you get the full attention of three people you chose by their portfolio, for three or four months straight. That's a rare luxury — even at the fee we ask.
It works only one way: a studio with this workload can't afford an "average" client. We choose projects the way clients choose us. And that's the relationship the three-a-year format makes possible.