A performance creative team should be organised around four decisions rather than four crafts: who owns the offer, who owns the brief, who owns the review, and who owns the read of results. Assign each decision to exactly one person, then buy or automate the production. Teams organised by craft (a strategist, two editors, a designer) hit a ceiling set by editing hours, and no amount of extra headcount in that shape produces more published ads per week.
The tell that a team is craft-shaped is a simple question: who decides which ads go live this week? If the answer is "we discuss it", you have four crafts and no owners, and your throughput is whatever survives the discussion.
The four decisions and who owns each
| Decision | What it actually is | Owner | Cadence | Failure when unowned |
|---|---|---|---|---|
| The offer | What the viewer gets, and the condition on it | Growth or marketing lead, with commercial authority | Monthly or quarterly | Ads default to "learn more", which produces creative nothing can rescue |
| The brief | Offer, audience objection, angle, constraints, format | Creative strategist | Weekly, one brief per test | Every output varies on everything, so the test compares nothing |
| The review | Publish, regenerate or cut, against the brief | One named reviewer per set | Weekly, one pass | Approval rounds multiply, keep rate falls, launches slip by days |
| The read | What the last batch means and what changes next | Whoever writes the next brief | Weekly | Winners run until frequency kills them, losers get re-tested by accident |
Two rules make this work. The brief and the read should sit with the same person, because a brief written by someone who did not read the results is a guess. And the review must have one accountable owner per set, not a committee, because sequential reviewers turn five hours of attention into three weeks of calendar time.
The offer
The offer decision is commercial, not creative, and it is the one most creative teams do not own and cannot escape. A 14-day trial with no card, 30 percent off until Sunday, free returns for a year: these change performance more than any hook you will write. If the person who owns the offer is not in the weekly loop, the creative team spends its time finding the ceiling of a weak offer.
Give this decision a monthly slot with commercial authority attached. Not a suggestion, a decision.
The brief
One person, one brief per test, five fields. The strategist role is defined by this artefact and by nothing else. If your strategist is also editing, the brief will be the thing that gets skipped, because editing has a deadline and briefing does not.
Volume: one strategist can write eight to twelve briefs a month across one or two accounts and still read results properly. Beyond that the briefs get generic, which shows up two weeks later as a keep rate drop.
The review
Publish, regenerate or cut, per output, against the brief and a written constraint list. Budget about five minutes per output. This is a checking role, not a taste role, and defining it that way is what keeps it fast.
Whoever owns the review should not have written the ad. That is the one separation worth preserving from the old agency structure.
The read
The read is the highest-skill decision on the list and the one most often delegated to a dashboard. It means deciding what the last batch proved, at what confidence, and what the next brief changes. Our creative testing workflow sets out the mechanics; the structural point is that this decision belongs to a person with a name, weekly, in a calendar slot.
Why an editor-shaped team caps output
Put the arithmetic on the table, because this is the part that gets argued about with adjectives.
An editor produces a short-form ad from existing footage in 30 to 90 minutes. Call it an hour. A full-time editor has maybe 25 productive editing hours a week after meetings, revisions and file handling, so about 100 produced ads a month at the optimistic end, and 40 to 60 is what teams actually see once revisions are counted.
Then apply keep rate, published divided by produced. At 60 percent, 60 produced ads is 36 published. That is the ceiling of one editor, and it is fixed.
| Team shape | Monthly produced | Keep rate | Published | Cost driver | What breaks first |
|---|---|---|---|---|---|
| 1 strategist, 1 editor | 50 | 60% | 30 | Editing hours | The editor's queue |
| 1 strategist, 2 editors | 100 | 55% | 55 | Editing hours plus coordination | Review, which is now one person's whole week |
| 1 strategist, 3 editors | 145 | 50% | 72 | Coordination overhead | Brief quality, because one strategist cannot feed three editors |
| 2 strategists, 1 reviewer, production bought or generated | 200 | 70% | 140 | Review time | Nothing, until offers run out of ideas |
The pattern in that table is the whole argument. Adding editors adds produced ads and lowers keep rate, because one strategist writing briefs for three editors writes thinner briefs, and thinner briefs fail review more often. You pay twice: once for the editor, once for the rejected work.
The fourth row is not a claim that tooling is magic. It is a claim about where the constraint sits. Move production out of the headcount and the bottleneck moves to review, and review is roughly five minutes per output rather than an hour, so the same team handles three to four times the volume. Keep rate is the metric that makes this visible, and it is the number to put on the weekly report before you argue about org design.
The strategist-to-account ratio
The number I would defend: one creative strategist per one to two active accounts, or per two to three accounts if the accounts share a product and an offer set.
That comes from the weekly workload rather than from a benchmark. One account at a weekly cadence needs one brief written (30 minutes), one batch reviewed (about an hour at five minutes per output), one read (45 minutes), plus launch and structure work. That is five to six hours a week per account of genuinely non-delegable thinking, and a strategist who is at four accounts is at 24 hours before anything goes wrong.
Signs the ratio is too thin:
- Briefs start reusing last month's angle with a new hook.
- The read becomes "the top ad by ROAS", which at a week's sample size is noise.
- Review turns into skimming, and keep rate climbs for the wrong reason.
- Nobody can say what this week's test is changing.
Agencies run this differently, because client management is a separate load. Our agency creative workflow covers the shape that works when account handling and creative strategy cannot be the same person.
What changes when production is automated
Three things change, and one does not.
The editor role changes rather than disappearing. What remains is the 10 to 20 percent of outputs that are nearly right and need a hand: a trim, a caption fix, a replaced shot. That is a genuinely useful job, and it is a smaller job than a full-time editing queue.
Review becomes the constraint. This is the main structural consequence. If you can produce 200 outputs a month, review at five minutes each is 16.7 hours, which is a real role with a real cost, and it needs an owner and a checklist rather than a shared inbox.
Brief quality becomes the highest-leverage work in the team. When production is nearly free, the difference between a 45 percent keep rate and a 75 percent keep rate is the difference between producing 89 ads and 54 to publish 40. At a loaded hourly cost of about EUR 45 (salary plus overhead, not just salary), that gap is worth more than an editor's salary. Our video ad cost calculator puts your own numbers through it.
What does not change: the offer and the read. No tool decides what to sell or what a result means. We build Genyad, so this needs saying plainly: there is no predicted performance score, nothing tells you which output will win, and there is no direct publishing to Meta or TikTok, so the launch and the read stay with your team. Those two decisions are where a performance creative team earns its budget, and they are the two you should staff for.
Frequently asked questions
What roles does a performance creative team need?
Four decisions need owners: the offer (commercial lead), the brief (creative strategist), the review (one named reviewer per set), and the read of results (whoever writes the next brief, usually the strategist). Production is a capability you can buy, generate or hire for, but the four decisions cannot be shared without throughput dropping. A team of two can cover all four if the offer owner sits nearby.
How many creative strategists per account?
One strategist per one to two active accounts at a weekly test cadence, or up to three accounts if they share a product and offer set. Each account needs roughly five to six hours a week of brief writing, batch review and result reading, so a strategist at four or more accounts starts recycling angles. The first symptom is briefs that reuse last month's angle with a new hook.
Why does adding editors not increase published ads?
Because it raises produced volume while lowering keep rate. One strategist writing briefs for three editors writes thinner briefs, thinner briefs fail review more often, and you pay for both the editing hours and the rejected work. Published output rises far less than headcount, and coordination overhead absorbs much of the rest.
Who should approve ad creative?
One named reviewer per output set, working from the brief and a written constraint list, with legal or brand signing off a standing constraints block quarterly rather than individual ads. The reviewer should not be the person who made the ad. Sequential multi-reviewer approval is what turns an afternoon of review into a three-week queue.