How much ad creative a Meta account needs each month
The number of new ads an account needs is not a guess. It follows from what you spend on testing, what a new ad needs before you can judge it, and how fast your winners tire. This guide sets out the arithmetic, then how to keep up with it.
By Dan BartleyUpdated 5 minute read
Key takeaways
- Meta’s delivery system now does much of the targeting, and the creative tells it who an ad is for.
- Work out the number from your testing budget divided by the spend one new ad needs before you can judge it.
- Launch on a steady schedule, so the next winner is proven before the current one tires.
- Count distinct concepts, not variations: a new crop of the same photo is not a new ad.
- Make the ads from one set of product frames, so volume does not depend on booking a shoot.
New creative is now the main lever in a Meta account
With broad targeting and Advantage+ campaigns, Meta decides who sees each ad. The ad itself tells the system who it is for: a different angle, scene or format reaches a different buyer. An account with a handful of ads gives the system little to work with.
That is why the same account can stall with the same budget and targeting: the ads stopped finding new people. More on this: how different do creatives need to be under Andromeda?
Work out how many new ads to launch from your testing budget
Divide what you can spend on testing new ads each month by what one new ad needs to spend before you can judge it. That gives you the most new ads the account can test properly. Launching more than that only spreads the budget too thin to read.
- Set a testing budget. A share of monthly spend kept for new ads, separate from the ads already working.
- Set the spend to judge one ad. A common rule is two to three times your target cost per purchase before you turn off an ad that has not converted.
- Divide the first by the second. That is the number of new ads the budget can judge each month.
| Step | Example |
|---|---|
| Monthly spend | $20,000 |
| Testing budget, at 20% of spend | $4,000 |
| Target cost per purchase | $50 |
| Spend to judge one ad, at two to three times the target | $100 to $150 |
| New ads the budget can judge each month | About 25 to 40 |
For high-priced products, where purchases are slow to arrive, judge early on add to carts or checkouts, then confirm on purchases. More detail: how many new creatives do we need per month at our spend level?
Replace winning ads before creative fatigue sets in
Every ad tires as the same people see it again. The signs show up together: frequency climbs, click-through rate falls and cost per purchase rises, on an ad that used to work.
- Frequency on the ad rising week on week
- Click-through rate falling while impressions hold
- Cost per purchase rising on the same ad and audience
- CPMs rising across an account with few new ads
Waiting for a winner to fall off leaves weeks with nothing proven to replace it. Launch new batches on a schedule instead, sized to the testing budget, so a replacement is already working when the current winner tires. More on this: what to do when creative fatigue pushes CPMs and CPA up.
Count distinct concepts, not variations of one ad
A concept is a different reason to buy, shown in a different way. A variation is the same ad with a new headline, crop or colour. Both have a place, but only new concepts give Meta something genuinely new to show to new people.
| New concepts | Variations of one concept |
|---|---|
| The throw on a deck in late afternoon: warmth outdoors | The same deck frame with three different headlines |
| The throw over an armchair in a reading corner: a quiet night in | The same reading corner cropped 9:16, 4:5 and 1:1 |
| A close-up of the knit: the quality of the yarn | The same close-up with and without a price |
Plan each batch on a grid of angles against the types of buyer you sell to, and fill the gaps. More: concept testing or iteration? and how to organise creative by angle, persona and format.
Mix static and video ads, made for each placement
Static ads and short videos reach people in different moods and placements, so a batch should carry both. Make each placement in its own ratio rather than cropping one master to fit.
| Placement | Ratio |
|---|---|
| Reels and Stories | 9:16, vertical |
| Facebook and Instagram feed | 4:5 |
Meta publishes the recommended ratio for every placement in its aspect ratio guidance. Keep the product and the message clear of the edges, where buttons and captions sit.
Keep up with the volume without a photoshoot every month
Most accounts run short of new ads because each batch waits on new photos. Break that link: make one set of finished frames for each product, then build every concept, format and placement from it.
- Brief. What is running, what has worked, and the products to push.
- Plan concepts. Distinct angles, each with a reason to buy and a test plan.
- Make. Statics and video in every placement ratio, from the product frames.
- Launch and read. Judge each ad on cost per purchase once it has spent enough.
- Repeat. What won and why shapes the next batch.
On one account, an e-commerce brand Dan Bartley ran creative for, working this way took new ads from under 5 a month to 50 to 100 a month, and the ads improved as the volume rose. That is one account’s result, shown to explain the approach, not a forecast for yours.
Read the results to plan the next batch
Cost per purchase and return on ad spend decide which ads won. The other metrics explain why, which is what the next batch needs.
| Metric | What it shows |
|---|---|
| Hook rate (3-second plays ÷ impressions) | Whether people stop scrolling |
| Hold rate (ThruPlays ÷ 3-second plays) | Whether they keep watching |
| Outbound click-through rate | Whether they leave for your store |
| Cost per purchase and ROAS | Whether the ad made money |
An ad that stops people but does not sell needs a better offer or landing page. An ad that sells but few people stop for needs a better opening. More: which creative metrics matter?
Questions people ask about this
How many new ads should a small account launch each month?
As many as the testing budget can judge properly. Divide what you can spend on testing by two to three times your target cost per purchase. For a small account that may be a handful of new concepts a month, launched on a steady schedule.
Read the full answerHow long should a new ad run before I judge it?
Until it has spent enough to show a result on your target metric. A common rule is two to three times your target cost per purchase, or sooner if its hook rate or click-through rate is far below your other ads.
Read the full answerDo new ads reset the learning phase?
They can. Meta treats adding a new ad to an ad set as a significant edit that sends that ad set back into learning. Launch each batch as its own ad set, or through Meta’s creative testing tool, and leave the winning ad set alone.
Read the full answerHow do brands make that many ads without a shoot every month?
By making one set of finished frames for each product and building every concept, format and placement from it, so a new batch does not wait on new photos.
Read the full answerStill have a question? Ask us
