Creative team structure: in-house or agency?

Every studio growth head eventually has this conversation with finance: why does the creative line cost what it costs, and would it cost less somewhere else. The honest answer is dull. A competent five-person in-house pod runs at roughly the same annual cost as a mid-tier agency retainer sized to deliver a comparable volume of finished creative. That pod is two motion or video editors, an art director, a concepting lead, a producer. So the economics rarely settle the argument on their own. What settles it is what each model does with the learning a creative test produces.

The cost comparison is closer than it looks

An agency retainer bundles account management and strategy input into one monthly fee alongside a rotating production bench, which looks cheaper line by line than five full-time salaries plus tooling and overhead. But an in-house pod doesn't bill for strategy separately, doesn't rotate its people between accounts, and doesn't need a briefing cycle to re-establish context it already carries. Once a studio produces enough creative volume to keep five people fully utilised, the two models land within a similar range on pure cost. Below that volume the agency is usually cheaper, because the studio isn't paying for idle capacity it can't fill.

Where ownership actually diverges

The real difference shows up after the campaign, not during it. An in-house pod accumulates institutional knowledge about which hooks worked for which cohort, why a concept fatigued faster in one market than in another, and what a particular creative director's instincts turn out to be worth over time. That knowledge stays with the studio as long as the pod's members do, and it compounds across every test that follows.

An agency accumulates the same knowledge, but it belongs to the agency's account history rather than the studio's, and it gets shared, deliberately or not, with the agency's other clients working similar genres. A studio that changes agencies, which happens more often than either side likes to admit, loses access to a body of learning it effectively paid to generate. That is the argument for in-house which finance conversations rarely surface on their own: creative learning is an asset, and only one of the two models lets the studio keep it on its own balance sheet.

Speed cuts the other way

Agencies win decisively on one axis: burst capacity. A studio facing a sudden testing requirement, whether that is a new market entry, a competitor's format shift or a network policy change forcing a creative refresh, can lean on an agency's bench without waiting out a hiring cycle. An in-house pod sized for steady-state production has no slack for a sudden doubling of demand. Hire to cover a temporary spike and the studio is overstaffed the moment the spike passes.

This is why the studios getting the best of both rarely pick one model outright.

The hybrid pattern that actually holds up

Most studios settle into the same shape. A small in-house pod, often three to five people, owns concepting plus creative strategy, and keeps the final say on what ships, while overflow production and burst testing volume route out to one or two agency partners on a flexible, non-exclusive basis. The in-house team keeps the learning log, which records what worked and why, and for whom. Agency partners execute against briefs the in-house team writes, which keeps strategic control and institutional memory inside the studio even when hands-on-keyboard production happens outside it.

This sits close to the discipline described in an earlier UA Ledger piece, Creative Operations for a Scaling Game Portfolio, which argued that the bottleneck in a growing creative operation is rarely raw production capacity but the process that turns one test's result into the next test's brief. That process has to live somewhere permanent. An agency retainer, however good, is not a permanent home for it, because the agency's incentive is billable output rather than a studio's long-run creative memory.

In practice the decision isn't in-house versus agency at all. It is which parts of the work are strategic enough that losing the memory of them would cost more than any short-term saving, and which parts are commodity execution that can be bought and swapped without losing anything that matters.

One further test separates the two cleanly: ask what happens if the relationship ends tomorrow, whether that means a key in-house hire leaving or an agency contract lapsing. If the answer is that the studio loses its record of what worked and why, that function belongs in-house whatever it costs. If the answer is that a new vendor could pick up the brief within a sprint or two, it was never strategic, and outsourcing it was probably safe all along.

Related archive reading

• UA team structure by studio size
• Playable ad agencies: when to hire one