How Many New Ads Per Month? A Forecast Formula
- New ads needed per month equals ads retired per month divided by the win rate of new ads.
- Ads retired per month equals active concepts times 30 divided by the days an ad stays effective.
- Count concepts, not files: ten resizes of one idea are one ad for forecasting purposes.
- Spend sets a ceiling on how many new ads you can read, so a formula result above the ceiling is a budget decision and not a production target.
- A lower win rate raises the volume you need faster than a shorter ad life does, so measure your own win rate before you plan.
- Rerun the forecast monthly with the last 90 days of data, because ad life and win rate move with the season and the audience.
- Omniconvert Commander generates launch-ready assets from customer data and recommends scale, hold, reduce or kill on each ad for you to approve.
How many new ads per month does a brand need? Most teams answer with a round number, such as 20, and then argue about whether it is enough. A better answer is a short calculation from four numbers you already hold: how many ads you run, how long they last, how many new ones win, and what you can afford to test.
Ask a media buyer why the target is 20 new ads a month and the honest answer is usually that 20 felt right. Omniconvert Commander (formerly Nexus) is the AI eCommerce growth engine for Shopify brands: it reads your store, ad and customer data, ranks growth opportunities by profit impact, and builds the campaigns to act on them, and you approve what goes live. The formula below needs only a spreadsheet.
The forecast stops at monthly volume. For the signal on a single ad, see how to predict ad fatigue and creative fatigue forecasting. For cadence rules, see how often to refresh ad creative.
How many new ads per month: the formula
The monthly new-ad requirement is defined as the number of new ad concepts you must launch in a month to keep your active set the same size. Here is the whole forecast in two lines.
Ads retired per month = active concepts × 30 ÷ days an ad stays effective.
New ads needed per month = ads retired per month ÷ win rate.
The first line is attrition. If every ad lasts 30 days, you lose the whole active set every month. The second line is yield. Most new concepts do not earn a place in the active set, so you must launch more than you lose.
A creative refresh rate is defined as the number of new concepts you launch per period to replace ads that have stopped earning their spend. The formula above is that rate, written per month. It gives you the volume. It does not tell you which ad to retire, because that is the per-ad signal that the pages linked above cover.
The four inputs and where to find them
| Input | What it means | Where to find it | The usual error |
|---|---|---|---|
| Active concepts | Distinct ideas that spent money in the last 30 days | Ad account, grouped by concept | Counting every size and copy variant as its own ad |
| Days an ad stays effective | Days from launch to crossing your cost line | Retired ads in your account history | Averaging in ads killed early for other reasons |
| Win rate | Share of new concepts that earn a place in the active set | Your launch log for the last 90 days | Using a rate from a blog post, not from your own launches |
| Spend level | Monthly test budget divided by the cost of one readable test | Finance plan and your cost per result | Planning launches the budget cannot read |
Use the middle value of ad life, not the average, because one ad that ran for a year pulls the average up and understates your churn. Take win rate from your own launches. A published rate belongs to someone else's offer and audience.
Spend level works as a ceiling and not as a term in the formula. One readable test costs your cost per result times the number of results you need before you trust a difference. That arithmetic sits in how often to refresh ad creative, and it is not repeated here.
Picture a Monday planning call. The report says 10 ads are live, so the team plans 10 new ones. Nobody mentions that those 10 rows are one concept resized across ten placements. When the concept tires, all ten rows die together, and the pipeline that looked a month deep is empty in a week. Count concepts, and the 10 live ads become 1.
A worked example of the forecast
The numbers in this example are illustrative. They are not benchmarks, and I have not sourced them from any dataset. Replace them with yours.
An illustrative brand has 8 active concepts. Each stays effective for about 30 days. One in three new concepts becomes a keeper.
- Ads retired per month: 8 × 30 ÷ 30 = 8.
- New ads needed per month: 8 ÷ (1/3) = 24.
- Spend ceiling: a 30,000 test budget, divided by 1,000 per readable test, gives 30.
24 is below 30, so the plan fits the budget. Now change one input. If ad life drops to 15 days, retirements double to 16, and the requirement rises to 48. That is above the ceiling of 30. The same brand now has a budget problem, not a production problem.
Change the win rate instead. If only one in five new concepts wins, the original 8 retirements need 40 launches. A weaker win rate hurts as much as a shorter ad life. This is why the brief you write matters as much as the number of ads you ship. See how to write ads from customer data in the AI ad copy generator guide, and how to build static ads from your catalog in the Shopify product ad generator guide.
Forecast your ad creative volume in six steps
- Count active concepts. List the ads that spent money in the last 30 days and group them by concept. Count each concept once, however many sizes or copy variants it has.
- Measure how long an ad stays effective. For ads you have retired, count the days from launch to the day each one crossed your cost line, and take the middle value.
- Measure your win rate. Divide the new concepts that earned a place in the active set by the new concepts you launched over the last 90 days.
- Calculate the new ads you need. Multiply active concepts by 30, divide by days of effective life, then divide that result by the win rate.
- Set the spend ceiling. Divide your monthly test budget by the cost of one readable test, and cap the formula result at that number.
- Fix the gap and rerun monthly. If the formula exceeds the ceiling, raise the budget, lengthen ad life or shrink the active set. Rerun the forecast each month with fresh data.
If you want a pillar view of what it takes to ship that volume, the section on going from 20 ads to 100 a month covers production. The forecast here does not.
What to do when the formula exceeds your ceiling
Raising the budget is the plainest lever and the one finance resists. Lengthening ad life means improving the offer behind the ad, or widening the audience so frequency climbs more slowly. Shrinking the active set means running fewer concepts with more spend each, which makes every test faster to read.
Ad creative volume is a means. Fewer concepts, tested properly, beat a large set that no one can judge. The discipline is the same one a structured testing programme runs on.
AliveCor used Omniconvert to run a structured A/B testing programme and achieved +21% conversion rate, +5% revenue per visitor, and 94% statistical relevance across their experiments. [Omniconvert, AliveCor case study]
Before you add ad volume, check the page the ads land on. A weak landing page makes every new ad look like a loser and drags the win rate down. Omniconvert Explore runs the A/B tests, with a 23.2% average conversion uplift across 70,000+ experiments [Omniconvert Explore, 70,000+ experiments]. That is an average across tests, not a forecast for your store.
Counting retirements and refilling the set
Omniconvert Commander connects to Shopify, GA4, Meta, Google Ads, Klaviyo and Gorgias, and reconciles against the platforms daily. It measures True Profit per ad and gives a scale, hold, reduce or kill recommendation on each one. The reduce and kill calls give you a clean count of retirements for the first line of the formula.
For the second line, Omniconvert Commander generates launch-ready assets from customer data, so the new concepts come from what your buyers say and not from a guess. You approve each ad before it launches to Meta. It does not tell you your win rate in advance. That comes from your own launches, one month at a time. For the full product, see Omniconvert Commander.
FAQ: how many new ads per month
What is the monthly new-ad requirement?
The monthly new-ad requirement is the number of new ad concepts you must launch in a month to keep your active set the same size. You get it by dividing the ads you lose each month by the share of new ads that become keepers.
What is a creative refresh rate?
A creative refresh rate is the number of new ad concepts you launch per period to replace ads that have stopped earning their spend. Expressed per month, it is the same number this formula returns.
How many new ads per month does a small brand need?
Three of your own numbers decide it, not brand size: active concepts, days an ad stays effective and win rate. A brand with few active concepts and a long ad life may need only a handful. Use the formula with your own data.
How do I find my win rate?
Take the last 90 days of launches. Count the new concepts that earned a place in your active set, then divide by the new concepts you launched. Ignore ads that never got enough spend to be read.
Does more ad creative volume always help?
No. Volume past what your budget can read gives you launches you cannot judge. Cap the formula at your spend ceiling and spend the extra effort on better briefs.
How does Omniconvert Commander help with ad creative volume?
Omniconvert Commander generates launch-ready assets from customer data and measures True Profit per ad. It recommends scale, hold, reduce or kill on each ad, which gives you the retirement count for the formula, and you approve what goes live.
Plan volume from losses
Start from what you lose, not from a round target. Count concepts, measure how long they last, measure how often a new one wins, and cap the answer at what your budget can read. Rerun it every month. The number you get is a plan you can defend, and a gap you can name.