How fast can we raise the budget on a winning campaign without losing efficiency?
Our answer
The answer depends on spend.
Raise an uncapped campaign's budget in steps of about 20 to 30%, two to three days apart. Stop when the return on the added spend drops below your target.
At $200K a month or more, scale with a ROAS goal or cost cap and a budget well above spend. Loosen the goal 5 to 10% a week, within the CAC you can afford.
Why this is the answer
3 reasons and our judgmentBudget steps of 20 to 30%, two to three days apart, can double a budget in about two weeks. Flighted's June 2026 guide raises budgets 20 to 30% every 2 to 3 days, taking a $500-a-day campaign to $1,000 a day.
Average returns hide what the last dollars cost. Wayflyer, an ecommerce lender, advises judging the return on the next $500 a day. In its skincare example, blended ROAS fell from 4.5 times to 2.4 times as daily spend tripled.
Meta gives no safe step size. Its help page says a change from $100 to $101 is unlikely to restart learning and one from $100 to $1,000 may.
Our judgment
- Steps two to three days apart follow Flighted over LearnyBox. LearnyBox's three to four days is written for course creators at $10 to $20 a day, while Flighted's guide is for consumer brands.
- When a step falls short, hold or step back to the last level that met target. This applies Wayflyer's rule of scaling only while marginal ROAS stays above break-even.
- Raising budgets in 20 to 30% steps suits uncapped campaigns only. Common Thread Collective (CTC) says the bid, not the budget, paces spend in capped ones.
- Read each step's return over the days since the step, not a single day. Meta says daily spend can run up to 75% over budget on some days and suggests reading campaign results over a full week.
Our judgment, built from Flighted, LearnyBox, Wayflyer, Common Thread Collective and Meta: our reading of these sources, not a finding from any one of them.
What the sources say
25 statements from 20 sourcesCharley Tichenor raises budget 5% on Monday, Wednesday and Friday only when a campaign is at least 10% under Target CPA. No results shared.
“Marginal scaling, the rule I use: +5% when you're at least 10% under target CPA. Monday, Wednesday, Friday. That's it.”
Marin Istvanic says many brands spend more on ads than their business can carry, and should pull back, fix the economics, then ramp gradually from what works. No data or step sizes.
“Win at a spend level you can actually manage, tighten up the economics, then ramp gradually from something that's already working.”
Northbeam's Bryan Bumgardner suggests raising spend about 20% at each BFCM check-in on ads beating target, plus an overnight rule resetting budgets 30% above the previous day's start. No data.
“we create a 12:01 am rule that sets the budget at 30% above the previous day's starting point.”
Segwise's own rules of thumb for Meta: 3 to 6 concepts per ad set, scale budgets 15 to 20% a day, and add a new concept when one takes 80% of spend for 14 days. It gives no supporting data.
Lorenzo Pravata raises budget 10 to 20% when yesterday beat target, waits when only the 3-day number does, and cuts 5 to 10% when both miss, reading results outside Meta.
“Raise budget 10-20% when yesterday beat target, wait when only the 3-day number does, cut 5-10% when both miss.”
Nick Theriot raises daily budget by 20% every day that the previous day came in below target CPA.
“all we do is literally increase daily budget by 20%, and then we'll just literally do that every day that yesterday was below that CPA.”
Meta says targeting, creative or optimization event changes and adding a new ad are always significant edits, while a budget change may or may not restart learning, depending on its size. No safe percentage is given.
“if you increase your budget from $100 to $101, that isn't likely to cause one or more ad sets to reenter the learning phase.”
Meta says it may spend up to 75% over a daily budget on some days but no more than 7 times the daily budget in a week, and advises reading results at the campaign level over a full week.
“On a weekly basis, we won't spend more than 7 times your daily budget.”
Aleksandar Hadzhiyski says it is fine to double your budget during Black Friday because demand is so high.
“During Black Friday, there's so much demand, there's so much happening that it's okay to double your budget.”
Jon Loomer says a budget increase of around 15% is a safe step that should keep volatility low when scaling.
“A safe increase that will hopefully result in minimal volatility will be around 15%.”
Eric Zaworski says a testing winner sometimes dies when it moves to a scaling campaign with an order of magnitude more daily budget, so scale in steps you can afford and watch each one.
Aazar Ali Shad says going from $30K to $150K a month raised CAC from about $30 to $35 while revenue grew five times: rising CAC at much higher spend is arithmetic, but rising CAC at flat spend is a warning.
Meta counts large budget or bid changes as significant edits that can restart learning, but it publishes no numeric threshold for how large a safe budget change is.
“Changes to targeting, placements, the optimisation event or bid strategy count as significant edits, as do adding a new ad, pausing for more than seven days and large budget or bid changes.”
Flighted suggests splitting the Q4 Meta budget 15-20% Nov 1 to 20, 40-50% Thanksgiving to Cyber Monday and 30-35% Dec 1 to 15, locking creative by Nov 15 and scaling in 20-30% steps. No sample.
“Scale in 20-30% increments, spaced at least 24-48 hours apart”
Mike Beckham says that on Meta the faster you try to grow, the less efficient your spending gets.
“the faster you try and grow, the less efficient you get, you lose efficiency as you try and go faster.”
A DTC Podcast guest says not to raise budgets 500% on day one going into November and December, and to ramp up gradually.
“You're going to want to gradually ramp up towards that.”
CTC promotes a paid service that moves consumer brands' Meta budgets many times a day, sometimes 3x to 10x within hours. It claims about 24% better results in a 37-brand backtest, with no control group.
“in some cases, you literally add zeros”
LearnyBox, which serves course creators with $10 to $20 daily budgets, advises raising daily budget 15 to 20 percent every three to four days. It says big jumps restart learning.
“Increase your daily budget by fifteen to twenty percent every three to four days”
CTC's analysis of 504 stores: about 70 to 75% of a month's Meta spend carries over on existing ads and 25 to 30% must come from new creative; brands scaling winners carry over 85 to 100% or more.
Wayflyer's illustrative examples show ROAS roughly halving when daily budgets tripled, such as 4.5x to 2.4x in two weeks; it advises raising budgets gradually and judging the next $500 a day.
CTC sets Meta budgets higher than it expects to spend and uses cost controls so the bid, not the budget, paces spend, accepting daily spend swings as Meta finds cheaper moments.
“Inflated budgets with cost controls allow for fluidity of market access.”
Nick Terrio says you cannot raise a daily budget tenfold overnight and expect to keep the same cost per result.
“you can't just 10 extra budget overnight and it'll hold the same cost per result goal.”
Sam Palero scales budgets in steps of 10 to 30%, then usually waits two to three days before the next change.
“I like to think of things in triggers of anywhere from 10 to 30%, most of the time... I would generally not make another adjustment for probably two to three days.”
Flighted, an agency, advises raising budgets 20 to 30% every 2 to 3 days, so $500 a day reaches $1,000 in about two weeks. Its warning that bigger jumps reset learning cites an article, not a test.
“Raise budgets by 20-30% every 2-3 days”
Josh Koth describes a brand whose ROAS began declining after it kept scaling budget but stopped launching new ads for a month.
“for one month, she stopped launching ads. She was still scaling her budget, but she stopped launching ads.”
In the full answer
For readers who sign in- What to do this week3 steps
- The numbers behind it3 figures from cited sources
- Where each source standsA chart placing each source on the question
- The strongest disagreementThe best case against the answer, from a cited source
- How sure we areWhat its confidence rests on, and what would change the answer