Unlimited design subscriptions promise unlimited requests. They don't tell you how unlimited gets constrained in practice. Here's the honest model.

Unlimited Queue, Limited Slots: The Only Honest Design SLA Model

Most design subscriptions advertise "unlimited requests." Read the fine print: unlimited requests go into a queue, and one is worked at a time. If you have 30 requests in queue and one active slot, the 30th request doesn't start for 30 days. That's not unlimited delivery — it's unlimited backlog.

What "Unlimited" Actually Means

Unlimited requests means you can submit as many tasks as you want. It's a queueing promise, not a delivery promise. Subscriptions with unlimited requests and one active slot are limited to one task at a time, regardless of how many you've submitted.

For teams that need two simultaneous campaigns, one active slot means the second campaign waits until the first tasks clear. If you're running four campaigns in parallel — Q2 push, a new market expansion, an enterprise sales motion, and a product launch — one slot means three of them are blocked.

The Honest Model: Capacity-Driven

The honest framing is: you have X active slots. Each slot handles one active request at a time. All other submissions queue behind live work.

This is transparent about what you're buying. You're not buying unlimited output — you're buying X simultaneous workstreams, each moving at a 48h rate for standard tasks.

For most teams at $1M–$5M ARR: one slot (Core) handles week-to-week GTM needs. For teams running multiple campaigns simultaneously (Series A+ with active demand gen and sales): two slots (Ops Base) handles parallel workstreams without blocking.

The SLA Connection

When you have a committed number of active slots, the 48h SLA means something. One slot, one active task, 48h to complete it — calculable. "Unlimited" with no defined concurrency makes SLA promises difficult: what's the SLA when you have 20 tasks in queue?

What This Means for Your Planning

When evaluating design services, ask:

• How many simultaneous tasks can be in progress at once?
• What's the SLA after a task enters the active queue (not after submission)?
• If I submit 10 tasks, what determines the order they're worked?

If you can't get clear answers, the capacity model is opaque, which means your campaign planning can't rely on it.

Sako publishes the capacity model: 1 slot for Core, 2 for Ops Base, 3+ for Enterprise →. Delivery is 48h from Ready →. Queue management is visual in the dashboard.