To scale Facebook ads without resetting the learning phase, raise the budget on ad sets that already sell above break-even in steps small enough that Meta does not treat them as a significant edit, and leave targeting, creative and the optimization event alone while you do it. Start the steps as demand for your product begins to rise, then step back down once it levels off.
The rest of this post covers what Meta counts as a significant edit, how it spends a larger budget, and how to time the whole thing.
What is the learning phase on Facebook ads?
The learning phase is the period when Meta's delivery system is still working out how an ad set will deliver and perform. It starts when you create a new ad or ad set, and again when you make a significant edit to an existing one. Ads Manager shows "Learning" in the Delivery column while it lasts.
Meta says an ad set usually leaves the learning phase after about 50 results in the week after its last significant edit. During learning, Meta says delivery is less stable and cost per result is usually higher. If an ad set cannot get enough results to exit, the status reads "Learning limited."
You can check where an ad set stands by adding the "Last significant edit" column in Ads Manager and comparing it with the results column.
Which edits restart the learning phase on Meta?
Meta lists these edits as significant, which means each one sends an ad set back into the learning phase:
- Any change to targeting
- Any change to ad creative
- Any change to the optimization event
- Adding a new ad to the ad set
- Changing the bid strategy
- Pausing the ad set for 7 days or longer (learning restarts when you unpause)
A second group may or may not count, depending on how big the change is: the budget amount, the ad set spending limit, and the bid control, cost per result goal or ROAS goal amount.
With Advantage+ campaign budget, the budget sits at the campaign level. Meta says changing that campaign budget or the campaign bid strategy can send several ad sets back into learning at once. Meta also says the way Advantage+ campaign budget moves money between ad sets does not restart learning.
Does increasing the budget reset the learning phase on Facebook ads?
A budget increase resets the learning phase on Facebook ads only when it is large enough to count as a significant edit. Meta gives a pair of examples: going from $100 to $101 is unlikely to restart learning, and going from $100 to $1,000 may.
Meta does not publish an exact percentage. Its advice is to use realistic budgets and avoid frequent budget changes, since those can push an ad set back into learning.
A practical reading of that: make each step small next to the current budget, wait until the ad set has stable results at the new level, and only then take the next step.
How do you raise a Facebook ad budget step by step?
You raise a Facebook ad budget step by step by changing only the budget, in modest increments, with a few days of stable results between each one. Everything else in the ad set stays the same while you scale.
A simple order:
- Confirm the ad set sells above break-even ROAS, the ROAS at which ad spend equals the gross profit it brings in.
- Check the ad set is out of learning and has run a while since its last significant edit.
- Raise the budget by a modest step.
- Wait for cost per result to settle at the new level.
- Repeat, or hold if a step pushes cost per result past what you can afford.
Save new creative for a separate ad set or for after the scaling ends. Adding a new ad to the ad set you are scaling is itself a significant edit.
How does Meta spend a daily budget after you raise it?
Meta treats a daily budget as a target over the week from Sunday to Saturday. On days with better opportunities it may spend up to 75% more than the daily budget, and across the week it will not spend more than 7 times the daily budget.
If you change the budget mid-week, Meta prorates spend for the rest of the week and adds a margin of up to 25% of the daily budget. So a raise on a Wednesday does not let Meta spend the old week's unused money on top.
Say your daily budget is $100 and you raise it to $120. Some days may come in above $120 and some below, and a full week at the new level stays at or under $840.
What is the difference between vertical and horizontal scaling on Meta?
Vertical scaling on Meta means raising the budget on an ad set that already works. Horizontal scaling means adding new ad sets, audiences or ads next to it.
Each approach has a learning cost. A large vertical jump can count as a significant edit. Every new ad set starts in the learning phase from zero, and Meta warns that many ads and ad sets split what the delivery system learns about each one.
| Approach | What it answers | What it misses |
|---|---|---|
| Raising the budget on a working ad set | How to buy more of what already sells | A large jump can restart learning |
| Adding new ad sets or audiences | Whether new groups of buyers respond | Each new ad set learns from scratch |
| Scaling on last week's results | Which ad sets worked recently | Spend tends to peak after demand does |
| Scaling into a demand window | When to start raising spend and when to stop | Weak creative or a slow landing page |
Raising budgets and adding ad sets are ways to add spend. Scaling on results and scaling into a window decide when you add it, and that timing matters as much as the method.
When should you scale Facebook ads?
You should scale Facebook ads as demand for your product starts to rise, so the budget steps are finished by the time buyer interest is at its highest. Scaling on last week's results puts spend on the same path a few weeks late.
A demand window is a stretch of time, with a start and a stop, when demand for a specific product is rising. What is a demand window covers the idea in depth. For scaling, what matters is that a window has a start you can plan the first step around and an end you can plan the last step around.
Windows come from build-ups that give buyers a reason to act:
- Home sales climbing for a third month, so new owners buy smart locks and window blinds in the weeks after they move.
- Pet adoptions rising for several months, so new owners shop for crates, beds and pet insurance.
- Preorders building ahead of a game console's release, so buyers want spare controllers and headsets ready for the first day.
Verdius finds each product's windows ahead of time with a start and a stop, ranks them against each other and lists the events behind each one. You still run the ads in Meta yourself. See how the windows are found.
How do I scale Facebook ads for a demand window without resetting the learning phase every time I raise the budget?
Start the steps a little before or right as the window opens, and change only the budget. Small steps with stable days in between keep each one below what Meta treats as a significant edit. Put new creative or new audiences in separate ad sets, or hold them until the window ends, so the ad set you are scaling keeps its learning. Verdius gives you the audience for each window by age, income, location and interests, which you can build as its own ad set.
Can you schedule a Meta budget increase for a known window?
Meta's budget scheduling lets you plan budget increases for set periods in advance, which suits a window with a known start and stop. You pick the period and raise the daily budget by an amount or a percentage. Verdius gives each window its start and stop dates, which you can enter as the scheduled period.
Meta sets these limits:
- It works only with daily budgets.
- You can add up to 50 high demand periods per campaign or ad set.
- A scheduled budget cannot go above 8 times the daily budget.
- Each period must last at least 3 hours.
When a period ends, the budget returns to the original daily budget on its own. That handles the step back down for you.
How do you scale Meta ads back down after a window closes?
You scale Meta ads back down by lowering the budget in steps, the same way you raised it, until you reach your base level. A large cut is as much a budget change as a large raise, and Meta judges both by size. Verdius sends an alert when the window closes, which gives you the day to start stepping down.
Lower the budget rather than pausing if you expect another window within a few weeks. Meta counts a pause of 7 days or longer as a significant edit, so a long pause means a fresh learning phase when you restart. When to pause ads covers that choice in more detail.
Then compare cost per result and ROAS inside the window with the weeks around it. The gap shows how strongly your product moves with the events behind the window, and it tells you how far to push the next one.
What tool tells you when to scale Meta ads?
Verdius is software that predicts when demand for a product will rise and who to target then. It tells you when to start the budget steps and when to bring them back down.
For each product it gives you:
- Optimal windows to run ads, each with a start and a stop, ranked against each other
- The events behind each window
- The audience to reach, by age, income, location and interests
- Alerts when a window opens and when it closes
Verdius reads Meta along with Google Ads, TikTok, Reddit, Snapchat, LinkedIn, Microsoft Advertising, ChatGPT Ads, Google Analytics and Search Console. It never launches, edits or pauses ads, so it can never restart learning. The pricing page shows what each plan covers.
So how do you scale Facebook ads?
Scale Facebook ads by raising the budget in modest steps on ad sets that already sell above break-even, with targeting, creative and the optimization event left alone. Begin as demand for the product starts to rise and step back down when it levels off. Keep the ad set running at a lower budget between windows so it holds on to what it has learned.