Scaling budget after a good week: step up 20% a day or double it once?

Picture a clinic whose Search campaign has hit its target CPA for three weeks, and the owner wants more leads before the holiday rush. The argument splits into the same two camps every time. One raises budget 20 percent every day or two and watches CPA closely. The other doubles it once and puts up with a few messy days while bidding settles. The step-up camp says it limits the damage if the good week was just luck. The double-once camp says small increases can keep Smart Bidding stuck in the same learning state for weeks, and that a daily budget cap often throttles delivery before the change even registers. Which way would you go, and does it change with conversion volume?

Check why the campaign is limited before picking a method. If it's hitting the daily cap by mid-afternoon, a 20% bump often just gets eaten by the same cap, so you learn nothing. If it's losing on rank or bid, more budget won't fix CPA. Volume matters too. With only a handful of conversions a week, I'd lean toward doubling once, since small steps can sit inside normal noise for weeks.

Is it actually capped by budget, or is that a guess?

Agreed on checking the cap first. The fastest way to settle it is the Budget lost impression share column, which shows how often the campaign missed auctions because of budget, instead of guessing from one afternoon. Pull at least two weeks, since a single busy day can mislead.

Keep in mind that Google can let daily spend run up to twice the daily budget on some days, as long as the monthly total stays in line. So a campaign that looks capped on one day may just be spiking. Average it over the month before you decide a 20% bump won't help.

Knowing the 2x daily allowance helps, but on social it's rarely what decides the outcome. On TikTok or Meta, a good week often rides on one or two creatives, and doubling the budget just pushes the same ad harder into the same people. Frequency climbs and CPA drifts up within days, so it looks like the scale-up failed when it was really ad fatigue. A safer rule of thumb is to step up 20 to 30 percent every few days while the creative is still holding, and only double if you already have fresh variants ready to absorb the extra spend.

Agreed on fatigue, and it's sharper for local services because the audience is so small. Take a dental clinic running Meta to a 10 km radius. That's maybe a few tens of thousands of people, so doubling the budget can push frequency up fast and the same faces see the ad every day. The early warning usually isn't CPA. It's CTR sliding while CPM stays flat, and that tends to show up a few days before CPA moves. On TikTok the equivalent is the 6-second view rate dropping on the same video. I'd make that the gate for the scale decision: if the hook metric is already declining, step up in smaller increments or hold until a new creative is live, whichever camp you're in.

The 6-second rate works fine as a gate, but on TikTok it often dips right after a budget bump. Delivery widens to colder viewers, so the drop doesn't necessarily mean the video is worn out. I'd compare that same video's rate before and after the change instead of against a blended weekly average. Are you splitting it that way?