How to choose a UA creative agency for your game

Choosing a UA creative agency for a mobile game comes down to four moves: match the agency type to your gap, check their portfolio against your genre rather than their showreel, run a small paid pilot before any retainer, and watch for the red flags that predict a bad handoff. The flags are familiar. Vague ownership of iteration; no named team; pricing that only makes sense at a volume you don't need yet.

That answer holds for any studio, but the detail changes depending on what is actually missing internally. A studio with a strong in-house video team but no playable capability needs a different agency than one with no creative function at all, and the rest of this piece is about telling those cases apart, then verifying with a small paid test before committing budget.

What types of UA creative agency exist?

The market splits into a handful of recognisable shapes, and most vendor conversations go faster once you know which shape you're talking to.

Full-service growth agencies run creative alongside media buying, and sometimes measurement as well, positioning themselves as an extension of your growth team rather than a supplier. They suit studios that want to outsource the whole function rather than production alone. They charge accordingly.

Creative production shops make video and static and motion work at volume without touching media spend, which suits studios that already run their own buying and just need more hands on the concept-to-variant pipeline.

Playable specialists build interactive ad units only, usually across several networks and SDKs at once. Their value sits in the technical craft a general video shop rarely has in-house: build size discipline, device compatibility, QA that catches a broken end card before a network does.

UGC shops produce creator-style and influencer-style ad content, sourcing talent or directing it to shoot in the format that now dominates a large share of mobile game creative. They're fast. They produce a lot of one specific look, and strategic depth isn't what you're buying.

Network in-house creative studios, offered by some of the larger ad platforms as part of a media relationship, produce assets at no separate line-item cost but usually only for campaigns running on that network, which limits their usefulness if you buy across several.

What should I look for in a creative agency?

Start with genre-relevant work, not overall reel quality. An agency that has never made a hook for a 4X strategy title will need a slower ramp on a 4X brief than one that has, regardless of how polished their general reel looks.

Ask who actually works on your account day to day, by name, not by department. Agencies sell against senior creative directors in the pitch and then staff junior producers on delivery; an agency that names the working team upfront and keeps them stable is a better sign than one that won't commit to names.

Check their iteration model. A good agency treats the first delivered batch as a starting hypothesis rather than a finished product, and it has a defined process for reading performance data back into the next round of concepts. Ask them to walk through how they used a client's fatigue data or hook-rate data to change a brief on a past project, described generally rather than by name if confidentiality applies. If the answer is a shrug, you have learned something.

Ask about capacity and exclusivity. Some agencies work with several games in the same genre at once; others deliberately cap category overlap. Neither is automatically wrong, but you should know which one you're getting and what it means for how fresh your concepts stay relative to competitors using the same shop.

What are the red flags?

Pricing that only works at a volume far above what you need is common and worth naming. An agency built around large monthly asset counts will often try to sell that same package to a smaller studio, and the unit economics rarely favour the smaller buyer.

Vague answers about revision rounds are another. A contract that doesn't specify how many rounds of feedback the fee covers, and what happens past that number, tends to produce billing disputes once a campaign is live and iteration needs to move faster than the contract assumed.

Watch for an agency that cannot describe a failure. Every agency with real client history has had a concept underperform or a playable get rejected by a network; one that claims a flawless record is either new or not being straight with you.

Be careful, too, of an agency that wants exclusive access to your performance data before any work has started. Sharing enough context to brief well is normal. Handing over raw dashboards before a pilot has proven the relationship is not.

How do I run a paid pilot before committing?

Treat the first engagement as a test, not a launch. A single concept sprint, scoped to a fixed number of variants at a fixed price, tells you more about an agency's real process than any pitch deck.

Give the pilot the same brief rigour you'd give an in-house creative team: a clear hook direction, the genre context, the audience context, and the performance metric you will judge the work against, whether that is hook rate or install cost, or a downstream retention signal if your test window allows it.

Judge the pilot on more than raw performance. How closely did the delivered work match the brief? How did the agency handle a mid-pilot change if you asked for one? Once results came in, how clearly did they explain their own read on what worked and what didn't? An agency that can only report a number, without a point of view on why it moved, isn't going to help you iterate once the retainer starts.

Only convert a pilot into an ongoing arrangement once you've seen a full cycle, brief through to a read on performance, rather than judging on delivered assets alone. The production is usually fine. The part that predicts whether the relationship holds up over a year is the iteration loop behind it.

Related archive reading

These articles provide related context and remain subject to their stated review status.

• Operating Brief: What Gamescom's UA Budget Signals Meant
• The Q4 creative calendar: when to lock, when to test