Performance creative, tested.

48 to 72 hours for first concept territories on a rush brief. Five to seven business days for a full launch batch on a standard brief, including statics, video cuts and platform-native variants. Always-on clients run on a weekly or biweekly drop schedule agreed in the retainer.

We track at the territory level, not the variant level, because that's how the algorithm sees it. Hook rate and 6-second view rate on video. Thumb-stop and CTR on static. Then CPA and ROAS by territory once spend matures. A territory that holds CPA for two weeks gets extended. A territory that breaks gets retired and replaced from the brief. We don't promise specific performance numbers up front, the variables outside our control (offer, audience, landing page, attribution model) move them too much. We do promise the production cadence, the entity-distinct discipline, and the testing matrix.

Project-based for one-off campaigns. Monthly retainer for always-on testing programs, billed quarterly in advance for retainer clients. The quote depends on channel mix, concept territory count and refresh cadence, which we lock after the first call. We share the full pricing once we understand the brief, because anything we quote before that is either too high or too low for the actual work.

Depends on the media spend level and the channel mix. A typical launch batch for a Meta-led campaign is three to five concept territories, with 20 to 40 statics and 6 to 12 video cuts per territory in the first drop. Always-on retainers usually ship dozens of entity-distinct creatives per month against a running test plan. We size the volume to what the algorithm is asking for at your spend level, not to a fixed package.

Brand guidelines if you have them. Product imagery, current best-performing ads, current losing ads, audience and offer notes, channel mix, media spend level, KPI targets. Access to your ad accounts if you want us to look at fatigue data directly. If brand guidelines don't exist, we build a working visual system in the first sprint and hand it over.

Art directors from agency backgrounds (think Publicis, Ogilvy, WPP) lead every concept territory. AI specialists handle production volume. Performance creative leads run the testing matrix. No junior teams running the brief alone. The creative direction layer is the single thing we won't compromise on, because it's where the AI slop risk lives.

Like you. We train our models on your existing brand assets, much in the spirit of Coca-Cola's StyleID approach, so the output sits inside your visual system rather than the model's default look. For brands without a documented system, we build one in the first sprint and hand it over. Avoiding the trust penalty is the single most important judgment call on every project, and it sits with the art director, not the model.

Every batch passes a brand-safety review before delivery, covering on-brand consistency, factual claims if any, and platform compliance. Where talent likeness is involved we secure rights documentation. We use commercially safe generation models and keep the workflow logged, so if an AI disclosure is required (the IAB's January 2026 framework is voluntary but increasingly expected on regulated channels) we can supply the metadata.

You own the final deliverables. We use commercially safe models and document the workflow. For the legal side in full, see our Copyright and AI guide .

Yes. Talent sourcing in the markets you sell into, voice-over and subtitle localization, regulatory copy variants by market, currency and offer logic. We've shipped campaigns across the US, EU, UK, Middle East, APAC and Greater China. Cross-border into Douyin and Xiaohongshu is a lane we run regularly for brands launching or relaunching in mainland China.

Native. Vertical first, sound-on hook design, creator-coded scripts, the cadence the platform actually rewards. We benchmark hook rate and 6-second view rate against TikTok Creative Center category norms before a batch ships, and design each hook to clear the 30 percent hook rate floor. Cross-posting Meta ads to TikTok is a known way to underperform.

Our production team is based in Shanghai with creative direction distributed across Europe and Asia. Most clients run on a 24-hour build cycle with that setup. Brief at end of day, first output at start of next day in their time zone. For Greater China work this is also where the in-platform fluency lives.

Fine. Many of our clients keep a brand agency for tentpole work and use us for the always-on performance creative the brand agency isn't built to staff at volume. We deliver into your naming convention and step back from the brand-strategy lane.

Two rounds of revisions are included in the standard concept quote. Beyond that we quote hourly, but it's rare. Inside the always-on retainer, revisions roll into the next cycle since the workflow is iterative by design. Rejected concepts stay yours, and we keep the source files on our side for 12 months in case you want to revive a direction later.

Yes. Send us one campaign brief. Inside one week you get a first concept batch and a refresh-cycle estimate sized to your media plan. No retainer, no commitment, the work is yours either way.