The 20% Rule in Action: How We 10x'd a Client's Spend in 12 Days Without Blowing It Up

The fastest way to scale a Facebook account is to be slow about it.

I know how that reads. But after years of doing this, I'm convinced the operators who scale the hardest are the ones who refuse to make big moves. Slow is fast, and fast is slow. Let me show you what I mean, because we recently took a client from roughly A$1,250 a day in spend to roughly A$12,500 a day in twelve days, and we did it by being deliberately boring.

That's a 10x. And there wasn't a single dramatic decision in it.

The three ways people try to scale, and why two of them hurt

When a brand wants to scale spend, there are really three roads people take. I've tested all of them. Two of them quietly work against you.

Campaign duplication. You've got something profitable at A$100 a day, so you duplicate it at A$1,000 a day and hope it holds. The problem is obvious once you've felt it. Just because an ad is profitable at A$100 a day does not mean it survives at A$1,000. You're putting far more stress on the system to find more buyers, and the ad might only work at the smaller number. So you jump up, it breaks, you drop back, you jump up again. It's a lot of lurching around the account, and the up-and-down is exhausting and inconsistent.

Cost caps. This one looks clever. You set a big budget and let the system only spend when it can hit a target cost per purchase. Your CPA stays lovely and stable. But here's what nobody tells you: it strangles your growth. To hold that cap, the system goes after the people most likely to convert, and those people are almost always your warm, bottom-of-funnel audience. Your frequency creeps up, you stop reaching genuinely new people, and the business stops growing. Great for a short-term efficiency win. Quietly terrible for long-term scale.

The 20% rule. You increase or decrease the daily budget by 20% based on whether you're hitting your target. That's the whole thing. The knock against it is that it looks slow. And on a tiny budget it is, 20% of A$10 a day is two dollars. But that "slow" is exactly why it wins, and the maths is the reason.

The compounding maths nobody sits down to do

Most people dismiss 20% because they're looking at one day in isolation. The trick is to look at the curve.

Picture you can lift the budget 20% most days because your creative, your product, and your landing page are all dialled in. Start at A$1,250 a day. Twenty percent up, day after day, looks like this:

• Day 1: ~A$1,500
• Day 3: ~A$2,160
• Day 6: ~A$3,730
• Day 9: ~A$6,450
• Day 12: ~A$11,100

Two weeks. That's nothing in the life of a business you're trying to build for years. And you've gone from a A$1,250-a-day account to one spending well over ten grand a day, without a single one of those reckless jumps that breaks performance.

To put the revenue side in perspective: at a 2.8x ROAS, A$1,250 a day is bringing in around A$105k a month. Hold those numbers as you climb to A$12,500 a day and you're suddenly looking at roughly A$1.05m a month. Same KPIs, ten times the size, twelve days of small, calm steps.

That's the case for boring. The compounding does the heavy lifting if you just let it.

The exact push and pull rules

Here's where most write-ups go vague, so I'll be precise. The whole system runs on one number and two timeframes.

Set your North Star first. Before you touch a budget, you need a target you live or die by. For us that's a new-customer CPA, the cost to acquire someone who's never bought before. I only watch new-customer CPA when scaling, because new customers are what actually grows the business. Existing buyers will purchase anyway through email and SMS, so I don't want them flattering the numbers. On the client I'm describing, the target sat around A$28.

To increase: look at the last 24 hours. If yesterday came in at or below your target CPA, increase the daily budget by 20%. That's it. You're not reading tea leaves. You're checking one number against one boundary and acting.

To decrease: give it 3 to 4 days first. This is the rule that saves accounts. When you have one bad day, do not touch anything. Sit on your hands. I'll allow three to four days of a high CPA before I cut spend, because the account very often auto-corrects on its own. The moment you start reacting to every bad day, you're trying to catch a falling knife, and your meddling makes it worse, not better.

That asymmetry is the heart of it. Quick to add on a good day. Slow and patient to cut on a bad one.

Set a hard deck. A hard deck is a floor you never scale below, whatever the performance, sized to the cash the brand can comfortably sit behind. It might be A$75 a day for a fresh account or several hundred a day for one with reserves. The hard deck keeps you in the game on the rough days so the creative has room to find its feet.

Surfing is for emergencies only. Doubling the budget two or three times in a day has its place, but that place is Black Friday or a tight, time-boxed event where you've only got days to spend. The other 99% of the year, it's just 20% up or 20% down.

What the 12 days actually looked like

So here's how it played out on this account, with the numbers rounded and anonymised but the moves exactly as we made them.

We came in with the creative already working. That matters, and I'll come back to it. The target new-customer CPA was about A$28, against an A$80 average order value, so the unit economics had real room. Spend was sitting near A$1,250 a day.

Day one, I checked the prior 24 hours. CPA was under target. Budget up 20%. Day two, same check, same result, up another 20%. By the third day we were past A$2,000 a day and the new-customer CPA hadn't moved.

Around day four we had a wobble, a single day where CPA popped above A$28. The instinct is to yank the budget down. We didn't. We left it completely alone, and by the next morning it had settled back under target on its own. That one held decision probably mattered more than any of the increases.

From there it was just rhythm. Check yesterday, hit target, nudge up 20%. We had two flat days in the middle where I simply held rather than pushed, because the prior day was borderline and there's no prize for forcing it. By the twelfth green-light day we were spending around A$12,500 a day, and the account was pulling close to 450 new customers a day instead of the 45-odd we started with.

A clean 10x. No duplication chaos, no cost-cap ceiling, no heroics.

The honest caveat

I have to be straight about the thing that made this possible, because without it the whole story falls apart.

The creative was already winning when we started scaling. The 20% rule is a scaling system, not a fixing system. If your CPA is sitting well above target, the answer is not to scale, it's to hold at your hard deck and go find a better creative. You scale a winner. You don't scale a problem and hope volume hides it.

When a brand is still hunting for that winning ad, this same account might crawl for weeks at the hard deck while we test, and only break out once a creative finally takes. That slow build is the unglamorous reality behind most of these curves.

So the contrarian claim holds, with a condition attached. Boring, rule-based scaling beats risky doubling almost every time. But it only works once you've earned the right to scale in the first place.

If you've got a genuine winner running right now, what's actually stopping you from being this patient with it?