When Should You Move a Winning Ad From Creative Testing to Scaling in Meta?

Short answer: Move an ad to Scaling when it has more than one sale at your goal on cost per customer, return on ad spend and order value together, confirmed by your third party attribution tool, over recent days. Before moving it, check that it is not already running in Scaling. After moving it, watch its delivery closely for about two weeks, because newly promoted ads often struggle to get budget next to established winners.

Promotion is the moment creative testing earns its keep. It is also where many accounts quietly lose winners. Ads get promoted on one lucky sale and flop. Real winners get called "ready" in a report and never actually move. Promoted ads sit in Scaling with no spend and nobody notices. This article walks through how we think about each step.

What does promoting to Scaling mean?

Promoting an ad means copying a proven creative testing winner into your Scaling Campaign, where it can compete for a much larger budget.

In our CTSG system, an account runs three kinds of campaign. Creative Testing finds winners with small, controlled budgets aimed at new people. Scaling is the main campaign where proven ads earn most of the budget. The Graveyard is a cost controlled home for ads that sell but not at a good enough cost. Promotion is the bridge from the first to the second.

This is called horizontal scaling: you put the winner into a separate campaign rather than only raising the budget on the test. Raising the test's budget in place is vertical scaling. Both have their place, and many winners get both. Horizontal is usually safer, because every campaign has a budget level where performance starts to drop, and you cannot know where it is until you hit it. A copy in Scaling does not put the original at risk.

What is the bar for promotion?

The bar is more than one sale, at goal on cost, return and order value together, confirmed by a source other than Meta, over a recent window, at the spend level the ad actually earned.

Each part of that has a reason.

  1. More than one sale. One sale can be luck. A single order can come from a friend, an existing fan or a coincidence. We want to see the result repeat before we bet real budget on it.
  2. Cost, return and order value together. Look at CPA (cost per acquisition, what you pay for one purchase), ROAS (return on ad spend) and AOV (average order value) as a set. Low priced products make CPA look good. High priced products make ROAS look good. A single large order can make everything look good. All three together tell the honest story.
  3. Confirmed by a second source. Meta's own numbers can be off in either direction. We judge on new customer numbers from a third party attribution tool. If Meta says it is a winner and the third party tool does not agree, the ad is close, not ready. If Meta misses the sales but the third party tool shows the ad clearly at goal, it is ready. See Meta under reporting.
  4. A recent window. Judge on the last several days, not lifetime. You want to know if the ad is selling now. An ad that sold well two weeks ago and has gone quiet is not a promotion candidate.
  5. At the spend it earned on. An ad that hit goal at a small daily spend has proven that much. Do not assume it will hold the same numbers at many times the budget. That is what the Scaling window is for.

Also check for the large order trap. If one order makes up a big share of the ad's revenue, judge it on CPA and on ROAS without that order. A pass that rests on one big basket is not a pass.

What should you check before promoting?

Check that the ad is not already in Scaling, the Graveyard or another campaign that receives promotions. It sounds obvious, but it is the most common promotion error we see.

Accounts with several Scaling Campaigns, a Retargeting campaign and a Graveyard make it easy to lose track. Someone promoted the ad last week. Or it was put into one Scaling Campaign but not another. Or it went to the Graveyard a while ago and nobody remembered.

The fix is a presence check every time:

  • Look in every campaign that receives promoted ads, one at a time.
  • Search by the exact ad name, including any unique code in it. Similar names are not a match.
  • Look over the ad's whole history, not just recent days, and count any delivery at all.

Then sort the result:

  • Not found anywhere: promote it.
  • Already in Scaling: it is not a promotion candidate. It belongs in your Scaling delivery review instead.
  • In the Graveyard: not a candidate.
  • Not sure the check ran: hold it until it does. Do not promote on memory.

Does daily frequency matter for promotion?

It matters for what happens next, but it should not block a promotion. We print it as a flag beside every promotion so the media buyer knows what to expect.

Daily frequency is how many times the average person saw the ad on a given day. In a creative testing campaign with strict exclusions, a daily frequency close to 1 means the ad is reaching people who have not seen your ads before. That is the ideal. An ad that sells at low frequency is creating demand among new people.

An ad that sold at higher frequency was more likely selling to people who had already seen your other ads. That is still a sale, and the ad may still deserve promotion. But here is what we have seen across accounts: promoted ads with higher testing frequency often get little or no delivery in a Scaling Campaign aimed at new people. They sell like remarketing ads, and Meta does not treat them as strong prospecting ads.

So we sort promotions by frequency first. Ads reaching mostly new people move first and have the best chance to get budget. Ads with higher frequency still move, but the buyer expects them to stall and watches closely. Some of those belong in a retargeting campaign instead. Always read frequency by day, never as a multi day average, because averages drift upward and hide the real pattern. More on that in why creative testing must run to cold audiences.

Why do promoted ads sometimes get no spend?

Because incumbents hold their spot. Established winners in a Scaling Campaign already have a long track record, and Meta keeps favoring them. A newcomer that crushed creative testing can still capture very little budget.

There are two failure patterns:

  • Starvation: the new ad barely spends at all. It was never really tested in Scaling.
  • The trap: the new ad gets budget but does not convert. It was tested, and it did not hold up at scale.

These are different problems. Starvation is not a fail. If an ad reaches the end of its Scaling window without getting enough delivery to judge, call it "never tested" and decide whether to restart or release it. Do not record it as a loss.

When several new promotions are starving at the same time, one remedy is a separate Ad Set for newcomers inside the Scaling Campaign, with a small spend minimum and a cap. That guarantees them a fair look without stealing much from the incumbents. We explain how in runoff Ad Sets and inside the Scaling Campaign.

How long should a promoted ad get in Scaling?

About two weeks, with early warning marks tied to the campaign's own numbers. Scaling moves on a slower clock than testing.

In the first day or two, you only learn whether the ad gets any spend at all. Slow starts are common and often turn around. So instead of fast kills, we use:

  • A benchmark CPA from the Scaling Campaign itself, based on its best recent performers.
  • A concern mark at a multiple of that benchmark spent with zero sales. Watch closely, look at adds to cart, and consider trimming.
  • A kill mark at a higher multiple with zero sales, or the end of the window, whichever comes first.
  • Graduation: when the ad lands a couple of sales at goal within its first week or so, it stops being "new" and gets judged like any established Scaling ad.

The dollar amounts come from each account's own benchmark, so they adjust automatically to cheap and expensive products alike. See kill thresholds for the reasoning.

What happens to the creative testing copy?

Usually it keeps running, and that is fine. We call this keeping the best version running.

When an ad is promoted, the original in creative testing often keeps selling. Several live copies of the same creative are normal. Judge each copy on its own numbers:

  • If both are at goal, keep both. Two profitable copies is a feature, not a conflict.
  • If the testing copy is failing while the Scaling copy works, pause the testing copy and note that the ad lives on in Scaling.
  • If the Scaling copy is not getting spend, that is a delivery problem to solve in Scaling. Do not pause a profitable testing copy because of it.

The usual pattern is that the creative testing version fatigues first. Recheck as performance shifts.

How do you make sure promotions actually happen?

Track every promotion call until the ad shows up in Scaling. A promotion called in a report but never done is a winner lost.

Two habits help:

  1. Keep a "promote still pending" list. Any ad called for promotion that the presence check still cannot find in Scaling stays on the list, with the date it was first called and its recent numbers, until it is moved, paused, or deliberately dropped. Do not let it disappear just because it stopped spending.
  2. Use a holding Ad Set. A single paused Ad Set in the testing campaign can hold every ad waiting to be promoted. It becomes a simple queue: the buyer opens it, copies the ads into Scaling, and marks them as moved. Nothing in it spends, because it stays paused.

At our agency, a person always makes the final move into Scaling. The system can find the candidates and organize the queue, but the promotion itself is a human call.

Key takeaways

  • Promote an ad when it has more than one sale at goal on CPA, ROAS and AOV together, confirmed by a third party source, over recent days.
  • Always run a presence check by exact ad name across every campaign that receives promotions before you promote.
  • Read daily frequency as a flag, not a gate. Low frequency ads have the best chance to get spend in Scaling.
  • Newly promoted ads get about a two week window in Scaling, with concern and kill marks based on the campaign's own benchmark.
  • An ad that never got delivery in Scaling was never tested. Do not count it as a fail.
  • Track every promotion call until the ad is actually in Scaling.

FAQ

How many sales does an ad need before scaling on Meta?More than one, at goal on cost, return and order value together, and confirmed by a source other than Meta. One sale can be luck. A repeat at goal is the first real sign the ad can carry more budget.

Should I scale by raising the budget or by duplicating into a Scaling Campaign?Duplicating into a Scaling Campaign (horizontal scaling) is usually safer, because it does not risk the original. Many winners also get careful budget increases in testing. Both can run at once.

Why is my promoted ad not spending in the Scaling Campaign?Established winners tend to keep the budget, and ads that sold at high daily frequency in testing often look like remarketing ads to Meta. A newcomer Ad Set with a small spend minimum can help ads that reach new people.

Should I turn off the creative testing ad after promoting it?Not automatically. If both copies are at goal, keep both. Pause the testing copy only when it is clearly failing on its own numbers.

How long should I wait before judging a promoted ad?About two weeks, with early warning marks based on multiples of the Scaling Campaign's benchmark CPA. An ad that graduates early with sales at goal moves to the normal Scaling rules.

Want to know if your winners are getting their shot?

Many accounts have strong tests that never made it to Scaling, or promoted ads that never got spend. Our free Meta account audit includes a full account audit plus net new reach and cumulative reach reports, so you can see where your winners stand. Request your free audit here.

Keep reading

More on the CTSG system:

See every article on the CTSG system

Ready for your free Meta audit?

We will review your account and run net new reach and cumulative reach reports, free.

Get your free audit

Check what you learned

Tap an answer to see if you got it right.

1. When should you promote an ad to Scaling?

Repeat sales at goal, confirmed by a second source, earn promotion.

2. What should you check before promoting?

A presence check stops you from promoting the same ad twice.

3. A promoted ad never got delivery in Scaling. Is it a fail?

No spend means no test. Don't count it as a fail.

Ready for the full test? Take the Scaling exam and get certified. Want this run on your account? Get my free audit.

About the author

Jake Mehani, founder of Creative Analytics

Jake has spent 15 years running Meta and paid social ads for consumer brands. Meta published his RBX Active Advantage+ case study, he taught the DMAnc Meta and TikTok ads masterclass, and he spoke at Motion's 2024 Creative Strategy Summit. His senior team runs media buying, creative and measurement like an in house team.

Jake on LinkedIn  |  Case studies  |  Free Meta audit

Previous
Previous

How to Tell if a Meta Ad Set Got a Real Test (Spread vs Concentrated Spend)

Next
Next

Why Is My Promoted Winner Not Spending in the Meta Scaling Campaign?