Why Did My Performance Max Campaign Suddenly Stop Spending?

Two problems look the same on a chart and they are opposites. Wasting spend means the money still goes out. It just goes to cheap clicks and products that never sell. This page is the other one. The money stopped going out at all.

So look at the spend line before you touch anything. If spend is roughly normal and only the sales fell, you have a waste problem. Our post "Fix Your Performance Max Campaign That Is Wasting Spend" covers those settings. If spend fell off a cliff, stay here and work the list below in order.

Start here. It is one field in the bid strategy, and it is the most common reason a campaign stops spending. The bid target you set (Target ROAS) tells Google what return you need per euro spent. Google only enters auctions where it thinks it can hit that number. Set it too high and Google skips nearly everything, so spend dries up. Nothing is broken. The campaign is doing what you told it.

Two quieter causes do the same thing. Thin conversion data, and targeting that is too narrow. Both leave Google with too few auctions it trusts.

The fix is to loosen the target and feed it cleaner conversion signal. We set the target around 20 percentage points below the ROAS the account is actually doing, so there is room to spend. Then we give it 2 to 4 days before judging the result.

A feed-only Performance Max campaign runs on your product feed. Products pulled out of the auction cannot spend. If your best sellers got disapproved overnight, the spend leaves with them.

Open Merchant Center, go to Products, then Diagnostics, and read what Google actually flagged. The usual cause is a price or stock mismatch between the feed and the site. Other culprits: a missing required attribute like brand, or an image Google rejected. Fix the real error, resync, and most clear within 24 to 48 hours. Do not request a review before you have read the error. ZenoX checks disapprovals on client accounts daily for exactly this reason, and Merchant Center health twice a week.

Wrong data poisons every decision after it. When the purchase signal stops arriving, Smart Bidding goes one of two ways. It overspends chasing bids it cannot check, or it sees no return and pulls back hard. On a sudden stop it is the second one.

Match the conversions Google Ads reports against the real orders in your Shopify admin for the same days. If the two do not line up, fix that before you touch a bid. ZenoX starts every performance drop with tracking and data accuracy, because fixing bids before fixing tracking is working backwards.

Open the change history and read the last two weeks. A bid strategy switch, a big budget jump, or heavy feed edits push the campaign back into learning. Spend goes patchy while it works itself out again. Small stuff like pausing one product usually does not do it.

Learning normally runs 1 to 2 weeks and needs around 30 to 50 conversions to fully exit. That is why we raise budgets 10 to 20 percent at a time. Then we give the account 2 to 4 days to settle before the next step. A smooth climb, not a spike followed by a crash.

Last check, not first. Seasonality moves spend without anyone touching the account. Q1 after the holidays is the weakest stretch for most ecom niches, and summer softens before Q4 picks back up. Fewer people search, so there are fewer auctions to buy.

It is the cheapest explanation to reach for and the most expensive one to assume. Only call it the season after the first four checks came back clean. Panic-changing a healthy account through a normal dip is how you actually damage it.

The order is the point. The bid target is the cheapest thing to check and the most likely cause, so it goes first. The season is the most expensive thing to assume, so it goes last. Work down the list and you find the cause without breaking anything on the way.

Do not raise the budget to force spend. The budget was never the ceiling here. A big jump also knocks a struggling campaign back into learning, on top of whatever is already wrong.

Do not change three things at once either. Then you cannot tell which one moved the number.

Change one thing. Wait 2 to 4 days. Then look again.