The 20% Rule Is Made Up: How to Decide When to Scale Meta Budgets Slow vs. Fast

It's 11pm, the house is quiet, and your finger is hovering over the daily budget field. The campaign's been sitting at a 2.8 ROAS for nine days straight. You type in a bigger number. Then you delete it. Then you type a smaller one. The 20% you've read about a hundred times. You nudge it up, close the laptop, and lie there wondering if you just left a month of growth on the table by being a coward.

I've watched founders do a version of that dance for years. So let me say the thing plainly: the 20% rule is made up. Nobody can tell you who invented it or why. It's not a Facebook policy, it's not in any documentation, and the algorithm has no idea you're obeying it.

It's a comfort blanket. And like most comfort blankets, it's doing a job, just not the one you think.

Where the 20% rule actually comes from

Here's the thing the rule is quietly protecting you from.

When you run at, say, A$100 a day, you buy a certain number of impressions. Picture them as dots. Some of those dots convert in a day, some take five days, a few take two weeks to come back and buy. So at any moment you're sitting on a pool of impressions that haven't paid off yet but will.

Push the budget up gently, 20% to A$120, and that pool grows by a sliver. The new impressions trickle in behind the old ones, the maths still looks healthy, and the account feels calm.

Now yank the budget from A$100 to A$500 overnight. You've suddenly flooded the account with a huge pool of fresh impressions that haven't had time to convert yet. For a few days the ROAS on paper looks ugly, because you're counting all that new spend against sales that simply haven't landed yet. The reported number tanks from 3 to maybe 0.8.

And what do most people do when they see 0.8? They panic. They pause. They tell themselves scaling killed the account.

That's the whole reason the rule exists. It's not optimising your growth. It's stopping nervous advertisers from spooking themselves into switching off a campaign that was actually fine. Facebook would rather you crept up and stayed in the game than leapt up, lost your nerve, and quit.

So the 20% rule isn't a scaling strategy. It's fear management dressed up as a scaling strategy. Useful if you're going to panic. Pointless if you're not.

I've watched accounts 10x overnight, and I've watched them die at 20%

Both of these are true at the same time, which is why the rule frustrates me.

I've seen accounts go from roughly A$10k a day to A$100k a day overnight and hold their numbers. Not gently, not 20% at a time. A clean 10x, in one move, on a day that was screaming to be scaled.

I've also seen accounts that should never have been touched get nudged up a polite 20% and fall apart, because the underlying signals were rotten and a bigger budget just poured petrol on a fire.

The size of the jump was never the deciding factor. What decided it was whether the account had earned the right to be scaled at all. The honest answer to "slow or fast" isn't a percentage. It's "what are the two numbers underneath telling me," and almost nobody is reading them.

Signal one: your target cost per acquisition is holding, with room to spare

This is the first thing I look at, and it's not your reported ROAS on its own.

Set your target. Say your break-even cost to acquire a new customer is A$50, and you want to be comfortably under it for sustainable growth, so your real working target is more like A$40. The question isn't "is the campaign profitable today." It's "is it sitting clearly inside my target with a buffer, and has it been there long enough that I trust it."

Here's the distinction that matters. If your target ROAS is 2.0 and the campaign is bumping along at exactly 2.0, you do not have permission to push hard. You're at the edge. Scale that and the first wobble tips you into a loss.

But if your target is 2.0 and you've been printing 3.0 for a week and a half, that gap is your runway. That's the account telling you there's slack in the system, that the spend has room to grow before efficiency drags back to break-even. A campaign beating its target by 50% for ten days has earned a real move, not a timid one.

So signal one isn't "am I profitable." It's "how much daylight is there between where I am and where I'd start losing money, and has that daylight been stable?" Wide and steady gap, you can be aggressive. Thin or jumpy gap, you cannot, no matter how tempting the line on the chart looks.

Signal two: your Shopify conversion rate, not just the Meta dashboard

This is the one most people skip, and it's the one that tells me whether a fast scale is safe or suicidal.

Meta will happily tell you a campaign is flying. But Meta is marking its own homework. The number I trust more sits in Shopify: are people actually converting on the site, and is that rate holding as more traffic arrives?

The check I run is simple. Pull up your Shopify sales-over-time view, ideally by the hour, and compare today against the same hours yesterday. If you've started pushing more budget through and the conversion rate is holding steady or climbing while the extra traffic floods in, that's a green light. The site is absorbing the new volume without choking. You can keep your foot down.

If the conversion rate starts sliding as volume climbs, stop. That's the site telling you the new traffic is worse quality, or you've pushed past the slice of the audience that was ready to buy. More budget from here just buys you more expensive tyre-kickers.

I'll give you the texture of a good day, with invented but realistic numbers. Imagine an apparel brand we're running. Early one morning the Shopify conversion rate is tracking visibly above the previous day's line, and it keeps climbing past 7am. So instead of letting it ride, we widen the gap on purpose. We double the spend. And nearly every hour after that, the gap gets wider, not narrower. By late evening the same hour that did A$450 the day before is doing closer to A$2,300. That's not a 20% day. That's a "the site is on fire in the good way, feed it" day, and the conversion rate gave us the confidence to do it.

The reverse happens too. Plenty of mornings the honest read is "conversion rate is soft today, leave it alone." Discipline is knowing the difference, and the difference lives in Shopify, not in Ads Manager.

So, slow or fast? Here's how I actually decide

Put the two signals together and the answer falls out on its own. You don't need the rule.

If your cost per acquisition is sitting well inside target with a steady buffer, and your Shopify conversion rate is holding or rising as you add traffic, you have permission to scale hard. That's when an overnight double, or more, is not reckless. It's responding to what the account is begging for. Creeping up 20% in that situation is the actual mistake, because you're rationing growth the account has already proven it can handle.

If either signal is shaky, the cost per acquisition is hovering right on the line, or the conversion rate dips the moment you add spend, then slow is correct. Not because 20% is magic, but because you don't yet have evidence the account can take more. Small steps here aren't timidity. They're you buying information cheaply before you bet big.

And if both signals are bad? Don't scale at all. A bigger budget never fixed a broken account. It just makes you lose money faster and more confidently. Fix the creative or the offer first, then come back to this question.

That's the whole framework. Two signals, read honestly, every time. "Slow is safe and fast is dangerous" is a story people tell themselves. The truth is that slow is right sometimes and fast is right sometimes, and the only way to know which is to look at the two numbers that actually carry the risk.

The percentage was never the point. The permission was.

Where to from here

If you want a gut-check before your next big push, run the two-signal test yourself tonight. Pull your trailing target-versus-actual on cost per acquisition, and put today's Shopify conversion rate next to yesterday's. If both are strong and steady, stop rationing and make the move. If they're not, you've just saved yourself a panicky 0.8 ROAS week.

And if you've been creeping up 20% for months and quietly suspect you've been leaving growth on the table, that's usually a sign the signals are stronger than your nerve. A fresh pair of eyes on your account, your real cost-to-acquire trend, and your site's conversion health will normally show you in an afternoon where you've been scaling too timidly and where you've been pushing into traffic that was never going to convert. That's exactly the read a Signal/Noise Audit is built to give you. Either way, I'd rather you decided on the numbers than on the nerves.

So, what's your account actually telling you to do tonight?