How a design and development subscription actually works
A flat monthly fee, one active request, pause whenever. Where that model genuinely beats hiring or an agency retainer, where it does not, and the failure modes nobody selling it will tell you about.
In short
- A subscription trades peak throughput for predictability. If you need five things at once, it is the wrong shape.
- The queue is the product. One active request is a constraint that protects quality, not a limitation to work around.
- Pausing is what makes it cheap over a year, and it is the feature most clients forget to use.
- For anything with a fixed scope and a deadline, a fixed quote is the honest answer, and we will say so.
Subscription design has been sold hard enough that the pitch has stopped meaning anything. Here is the mechanical version: what actually happens day to day, what it costs, and the three situations where we tell people not to buy it.
The mechanics
You pay a flat monthly fee. You add requests to a queue. We work the queue top down, one active request at a time, every business day, and you always know what we are on. When something is finished it goes live and the next thing starts. You can pause the subscription and only pay for the days you used, and you can cancel outright without a conversation about contracts.
That is genuinely the whole model. Everything interesting about it is in the consequences.
What it costs, and against what
Our subscription is 1,602 euro a month, and the PRO plan with two concurrent requests is 2,725. Those are the real numbers, they are on the pricing page, and they are what Stripe charges.
The honest comparison is not against a freelancer's day rate. It is against the three things a subscription is actually competing with:
- Hiring in-house. A senior designer or developer in Western Europe is a salary plus employer contributions plus equipment plus recruitment plus the months before they are productive. The real problem is not the money, it is that the workload is rarely constant enough to keep one person usefully busy, and you cannot pause a person.
- An agency retainer. Usually a block of hours, and hours are a bad unit. You end up paying for meetings, for scoping, and for the agency's estimation risk, which they price in because they have to.
- Project by project. Lowest commitment, highest friction. Every piece of work needs a scope, a quote and a negotiation before anything starts, which is fine annually and miserable weekly.
A subscription wins when your work is continuous and lumpy: a steady stream of things that each take days rather than months, arriving unpredictably. It loses when your work is one big defined thing, and we will tell you that rather than take the money.
Pausing is the actual feature
The pause is what makes the annual number reasonable, and it is the thing clients most often forget to use. If you have eleven days of work in a month, use the subscription for eleven days and pause it. The remaining days stay on the shelf until you need them.
Nobody selling subscriptions is incentivised to remind you of this, which is exactly why it is worth saying in public. A client who pauses correctly and comes back is worth more than one who quietly resents an idle month and leaves.
Where it breaks
You need five things at once
A queue is sequential by design. If you are launching in six weeks and need a brand, a site, an app and a campaign in parallel, a subscription is the wrong shape and no amount of goodwill fixes that. This is what the fixed quote route exists for.
The work needs deep, ongoing product context
Some work only makes sense with months of accumulated context: a complex product's internals, a regulated domain, a codebase with a lot of history. A queue that switches subject every few days does not build that. When we take that kind of work it runs as a scoped engagement, not as queue items.
Nobody on your side owns the queue
This is the most common real failure and it is the least discussed. The model needs one person on the client side who decides what is next. Without that, the queue fills with contradictory requests from different departments, priorities change mid-task, and the throughput everyone was promised evaporates. It is not a design problem or a development problem. It is an ownership problem, and it is worth solving before you subscribe.
The other door
We run scoped builds too, on a single fixed price agreed up front, and we operate them after launch. That is how our Reddit games are run, and it is the right answer whenever the work has a definable end. If you come to us with something that clearly wants a fixed quote, you will get a fixed quote.
A studio that only has one commercial shape will fit every problem into it. Having two means we can be straight with you about which one you are actually looking at.