A campaign finally works, so the instinct is to pour money into it. Double the budget, double the leads. It almost never lands that way, and the reason has nothing to do with the algorithm getting confused.
Why doubling breaks it
At its old spend level, the campaign could afford to be picky. It took the cheapest conversions available to it and ignored the rest. Double the budget and it still has to spend it, and the only way to do that is to move down into the auctions and queries it was previously skipping. Those cost more and convert worse.
So your cost per acquisition does not rise because something broke. It rises because you asked the system to buy the next-best inventory. Same machine, worse raw material. Worth being precise here: this is our read from running accounts, not a documented Google mechanic. What Google does document is that budget changes are not on its list of learning-period triggers, which rules out the explanation most people reach for first.
The rule we use
Increase by 20 to 30 percent. Never 2x. Then leave it alone for about ten working days before touching it again.
The waiting is not superstition. Google says a bid strategy can take up to three weeks, or one to two conversion cycles, to calibrate through the learning period. Change the budget every few days and you are stacking adjustments onto a system that has not finished absorbing the last one, and you will never know which change caused what.
Ten working days is a floor, not a law. If the account converts a handful of times a month, or the sales cycle runs long, you need more. Volume decides how fast you are allowed to move.
Read the result at the new spend
This is where most scaling decisions go wrong. People compare the new CPA to the old CPA and panic.
- Higher CPA, much higher volume. Often the right trade. You bought growth at a worse unit rate, which is a business decision, not a mistake, as long as the new rate still clears your pipeline economics.
- Higher CPA, flat volume. You moved too fast. Step back down and go again in smaller increments.
- Flat CPA, higher volume. Keep going. You have not found the ceiling yet.
And judge it against CAC and pipeline, not just the platform's cost per conversion. A cheaper lead that never becomes an opportunity is not a win.
One change coming
Google has announced that from 17 August 2026, campaigns using bid targets such as target CPA or target ROAS will behave more consistently when they are limited by budget, including after budget adjustments. If you run target-based strategies, expect some temporary fluctuation around that change and avoid reading it as the result of your own scaling.
The takeaways
- Raise budgets 20 to 30 percent at a time. Never double a working campaign.
- Wait about ten working days, longer if conversion volume is low or the sales cycle is long.
- Judge CPA at the new spend level: higher CPA at much higher volume can still be the right trade.
- Higher CPA at flat volume means you moved too fast. Step back down and go again smaller.