Change your budget without breaking delivery
Double a budget overnight and you start from zero. Raise it in steps and you keep what the campaign learned.
Why budget is sensitive
Budget doesn’t just fill a bucket. It changes how the algorithm buys. With more money, it has to find more people, so it reaches profiles it rated less promising. With less, it narrows down abruptly.
A big change is treated as a new context: the learning phase restarts, and cost per result rises while it lasts.
What counts isn’t the amount itself but the relative jump, and how often you change it. Two small increases on the same day disrupt as much as one big one: the algorithm never gets time to settle in between.
Scale up in steps
Increase in small steps, and let delivery settle before the next one. It feels slow when a campaign is working and you want to go faster. It’s still the fastest route to a high, stable budget.
The shortcut (tripling overnight) sometimes works. When it doesn’t, you lose both what the campaign learned and the money spent relearning it.
The alternative: duplicate
For a big jump, another option is to leave the current ad set as is and create a second one with the target budget. The original keeps its learning, the new one learns on its own, and you compare.
It costs more short term, since two ad sets learn at the same time. But it puts nothing that already works at risk.
Scale down gently, too
People forget this: cutting a budget in half disrupts as much as doubling it. If you need to ease off (end of season, out of stock), do it in several steps, not all at once.
Adwize suggests the steps and explains each one. Every change stays a suggestion until you approve it in the conversation, and you can set, in plain English, the threshold where you want a heads-up.