Design capacity: subscribe, hire, or both

Frame it as a capacity decision

“Should we hire a designer?” is a headcount question. The question that matters is how much design capacity the roadmap needs, how predictable that demand is, and how fast you need it.

Before choosing a model, I ask leadership teams to forecast design demand for the next two quarters, initiative by initiative: the onboarding rebuild, the pricing and packaging change, the enterprise admin surface, the fundraise narrative, the brand refresh. Size each one and note where it sits on the roadmap.

The shape of that forecast usually makes the decision for you. Steady, single-product demand that compounds over time describes a role. Demand that spikes around launches, raises and campaigns, and spans product and marketing, describes a queue. Staffing a queue with permanent headcount means paying for idle capacity in the troughs and still running short at the peaks.

Price in the full cost of a hire

Salary is the visible line. The bigger costs are time to value and management attention.

A senior design hire carries a recruiting cycle, benefits, tooling and a ramp period before the work changes your metrics. While the seat is open, the roadmap either waits or ships without design, and the second option creates debt you pay back later in rework and inconsistency.

There is also a retention risk leaders tend to underweight: a strong designer with half a job leaves. If you cannot offer a full, growing scope of ownership, the hire is fragile no matter how good the person is.

Hire when all three are true:

• There is sustained demand for at least one full-time designer over the next year.
• You need design judgment present in daily product decisions, not just in deliverables.
• You can give that person real ownership of a product area and the outcomes tied to it.

Where a subscription is the better instrument

A subscription converts design from a fixed cost into flexible capacity you can scale with the roadmap.

At ProductBloom the model is deliberately simple: requests go into one queue, senior product designers and a senior developer work it one task at a time, and finished work comes back within a matter of days. When priorities shift, you reorder the queue. When the roadmap quiets down, you pause.

It is the stronger choice when:

• You need to compress time to value, for example ahead of a launch or a raise, and cannot wait out a hiring cycle.
• The work spans disciplines: product UI, design systems, marketing site, brand and motion.
• You want senior judgment without adding a management layer.
• You need to validate direction before committing to permanent headcount.

Be precise internally about what “unlimited” means. It means nobody has to ration requests or negotiate scope per project. It does not mean parallel work. Sequencing is the point: one task at a time is how quality stays high and priorities stay honest.

The hybrid most teams should consider

The decision is rarely either/or. The best sequence is often subscription first, hire second.

Use subscription capacity to establish the foundations: the design system, the core flows, the marketing site. Then hire into a defined role with a clean system to inherit. Your first in-house designer starts on day one with tokens, components and patterns instead of an archaeology project, and you hire for the scope you actually have rather than the one you guessed.

Once a team is in place, the same capacity becomes overflow for peaks and specialist work, so in-house designers stay focused on the product areas they own.

Running it well

Whatever model you choose, the return depends on how the work is fed and governed.

1. Brief outcomes, not outputs. “Activation stalls at workspace setup” gets a better answer than “redesign the setup screen”.
2. Give designers the context product managers have: analytics, support themes, sales objections, constraints.
3. Name one decision maker per request and consolidate feedback into a single pass.
4. Prioritize the queue by business impact and review it on a fixed cadence with product and engineering.
5. Close the loop. Tie shipped design to the metric it was meant to move, and feed what you learn back into the queue.

If you are weighing the options, a free 30-minute call is enough to pressure-test your forecast. If a subscription is not the right instrument for your situation, I will say so.