How many days do you leave a new account alone before you touch the budget

I have doubt and i think is the most expensive doubt i have.

Every account i lose, i lose it in the week when i start to scale. Never in the boring weeks. So clearly i am doing something wrong at this moment and i want to know what the number is.

Actually people say "warm it up" but nobody say for how many days and nobody say how much you can raise. My last one i wait seven days, then i double the budget, dead in two days. Before this one i wait three days and go 5x, dead the same day, which ok that one is my fault.

So: how many days, and then what percentage per day. Somebody who runs many accounts must have a rule.

there's no number. there's a billing cycle. scale after the second clean charge, not after a calendar

grindstone's answer is the better version of the question and I'd build on it.

What I run, and it's a rule of thumb rather than a law: 1) no change of any kind until two billing charges have cleared, which at a low threshold might be three days and at a high one might be two weeks. 2) after that, increases of no more than 30% per day, and never twice in the same day. 3) no budget change on the same day as any other change — not creative, not audience, not payment, not profile.

Point 3 is the one people skip and I think it's the one that matters most. A 30% budget rise on a quiet account is a normal thing an advertiser does. A 30% rise on the same day as three new creatives is a different-looking event entirely.

I logged this for a year because I got tired of the folklore. 34 accounts, one platform, 2023 into early 2024.

Days before first budget increase	n	Survived 30 days
0-2	7	1
3-6	11	4
7-13	9	6
14+	7	6

And separately, by size of the first increase:

First increase	n	Survived 30 days
Under 50%	15	11
50-150%	12	5
Over 150%	7	1

Both curves point the same way and neither is clean, because the two variables correlate — the people who wait longer also tend to raise more gently, including me.

the second table is so useful. under 50% is basically double the survival of over 150%, that's not subtle at all

under 50%. that's the whole thread and it's in row one

34 account on one platform is more data than anybody post here in a year. Anyway your two variable are tangled, you say it yourself, but tangled data is still better than nothing.

Had a stretch in 2022 where I lost six in a row in the scaling week and eventually worked out it wasn't the scaling at all - it was that scaling is when I started running the offer that actually made money, and that offer was the problem.

Everything before that was a soft creative on a soft angle. The account survived it fine. The week I switched to the real one, that was the event.

Worth asking yourself whether your scaling week is also your switching week. Is it?

Ah. Yes. Actually yes.

I always test with a soft version first because is safer, and then when the numbers look ok i put the real lander and raise budget the same day. Every time. I never see it as two things.

So maybe my problem is not the days at all, is that i change the offer and the money in one afternoon and call it scaling. This is a very annoying thing to learn from a question about numbers, my friend, but thank you.