Free Resource
How Much Ad Creative Do You Actually Need?
Creative volume, creative diversity and creative velocity, translated from platform vocabulary into a plan a brand can scope.
Planning guide Written by the studio from its own practice; where it states a view, the view is the studio's. Figures quoted from a platform are attributed to that platform's documentation under Sources and should be verified at upload.
Every paid social team hears the same instruction from the platforms, the agencies and the tool vendors: you need more creative. The instruction is half right and unhelpfully vague. An account does need enough distinct creative to learn from and to refresh with, but "more" is not a number, and the number is not the same for any two accounts. This guide replaces the word with a method: what actually sets the size of a batch, which kind of volume matters, how to tell diversity from velocity, and how to write the result into a scope a studio can produce against. It makes no claim about what any batch will do, and it names no universal figure; the number comes from the account, not from this page. It is the planning behind performance creative produced as test batches.
What's included
- Count concepts, not files
- Size the batch from the questions, the delivery and the placements
- Diversity is how different the ideas are; velocity is how quickly a tired one is replaced
- Refresh from the account's reads, not from a fixed interval
- Fix the brief, references and approvals before blaming production
- Write the batch as a grid the account can read
Where the pressure for volume comes from
Three places, and they are worth separating. Delivery systems on Meta and TikTok distribute spend across the creative they are given, and their own guidance asks advertisers to supply several distinct pieces rather than one, so that the system has something to choose between. Fatigue is real: a piece an audience has seen repeatedly stops earning attention, and the only remedy is another piece. And testing needs candidates: an account cannot learn which opening, angle or format works from a single ad. Each pressure asks for a different kind of creative, which is why a single volume figure answers none of them.
Volume is the wrong unit
A batch of ads that differ only in caption colour is one ad. The unit that matters is the concept: a distinct persuasive idea, with its own opening, its own argument and its own proof. Variations sit inside a concept and change one element of it, which is what makes a test readable. An account that counts assets will overproduce cosmetic variants and underproduce ideas; an account that counts concepts knows what it is buying. When a platform or a vendor says diversity, it means concepts. When it says velocity, it means how quickly a new concept can replace a tired one. Neither is a count of files.
- Creative volume
- The number of distinct pieces of creative an account has available to run in a period. Meaningful only when it is broken into concepts and the variations inside them.
- Creative diversity
- How different the concepts are from one another in opening, argument, format and proof. The property the delivery systems' own guidance asks for; a large batch of near-identical pieces has none of it.
- Creative velocity
- How quickly a new concept can be briefed, produced, approved and live once the account's reads say a piece is tired. A property of the whole loop, from brief to launch, not of production alone.
- Creative bottleneck
- The step in that loop that sets its pace. In most accounts it is a brief, an approval or a claims review rather than the making of the frames.
Three questions that size the batch
The size of a batch follows from three inputs the account already has. What questions does the account need answered next, and how many concepts does it take to answer them? How much delivery can the account give each piece before a difference means anything, given its spend, its audience and its placements? And how many placements does each concept have to be cut for? Answer those and the batch sizes itself: concepts down, variations across, one cell per deliverable. A number chosen before those questions are asked is a guess dressed as a plan.
| Input | What it decides | Where it comes from |
|---|---|---|
| The questions the account needs answered | How many concepts the batch needs, and which variable each one tests | The account's own history: what has run, what was learned, what was never tried |
| Spend, audience size and delivery | How many pieces can be read against a control before the reads mean anything | The media plan and the platform's delivery on the account |
| Placements and formats | How many cuts and ratios each concept needs | The plan's placements, checked against the platform's current specifications |
| The refresh the reads call for | Which concepts are retired and which get a successor | Frequency, attention and response on the pieces already running |
| Approval and claims capacity | How quickly a new concept can be cleared to run | Whoever signs off on claims, product accuracy and brand fit |
Refresh from the reads, not from a calendar
A fixed refresh interval is a volume figure by another name. The honest trigger for a new concept is the account's own evidence: rising frequency, falling attention in the first seconds, a response that has drifted from the account's baseline on a piece that used to hold it. Those reads say which concept is tired and, often, why, which is the brief for its successor. A studio that can see the reads through read-only, revocable reporting access can build the next batch from them; one that cannot is guessing on the brand's behalf.
The bottleneck is usually not the production
When a team says it cannot produce enough creative, the delay is often upstream of production: a brief that takes weeks to agree, references nobody has collected, a claims review with no owner, an approval queue with one approver. Generative production changes the cost and the logistics of making frames; it does not change any of those. A scope that names who briefs, who supplies references, who approves claims and who signs off, with a date for each, removes more of the bottleneck than any production method does.
How to write volume into a scope
Write the batch as a grid rather than a figure: the concepts, the variable each one tests, the variations inside it, and the placements each variation is cut for. Add the naming convention the account will use to read the results, the decision rule agreed before launch, and the reporting access the studio has, if any. That document is the scope, the launch plan and the brief for the next batch at once, and it is what Unreal writes before any production starts. The structure is set out in the creative testing framework.
What this guide does not tell you
It does not give a number of ads to make, because there is not one. It does not say that more creative produces better results; it says that the right creative, in a batch the account can read, produces a decision. And it makes no claim about the results of any batch Unreal has produced: the studio publishes no performance figures and never guarantees advertising performance.
Sources
- Creative best practices for performance ads, TikTok for Business.
- Ads Guide: placements, sizes and specifications, Meta.
In practice
The service this guide supports: performance creative produced as test batches.
Frequently asked questions
How much ad creative do I need for Meta or TikTok?
Enough distinct concepts to answer the questions the account has, each cut for the placements it will run in, in a batch the account can read against a control. That depends on spend, audience and placements, so it is agreed per account rather than taken from a benchmark. Both platforms' own guidance asks for several distinct pieces rather than one.
What is creative diversity?
How different the concepts in an account are from one another: opening, argument, format and proof. A batch of one idea in a run of cosmetic variants has volume and no diversity.
What is creative velocity?
How quickly a new concept can go from brief to live once the reads say a running piece is tired. It is a property of the whole loop, including approvals, not of production speed alone.
How often should ad creative be refreshed?
When the account's own reads say a piece is tired: rising frequency, falling attention in the opening seconds, a response drifting from baseline. A fixed interval ignores the evidence and either wastes creative or leaves a tired piece running.
Does more creative mean better results?
No. More creative means more to read and more to pay for. The right concepts in a batch the account can read produce a decision; whether that decision improves results depends on the campaign, the offer and the media plan, and no production partner can promise it.
Start with one scoped batch.
Book a free 15-minute discovery call. We'll map the exact plan for your brand — no pitch deck, no fluff.