Target ROAS in Google Ads: how to set it without starving the campaign
Set target ROAS at or slightly below the ROAS a campaign has actually delivered over the last 30 days, and keep it above your break-even ROAS, which is 1 divided by your gross margin. A target above what the campaign already achieves makes Google bid lower and sit out auctions, so traffic shrinks long before efficiency improves.
Break-even ROAS = 1 ÷ gross margin
A store keeping 60 cents of each sales dollar after product cost, shipping and payment fees breaks even at 1 ÷ 0.60 = 167%. Below it, every ad sale loses money.
Achieved ROAS = conversion value ÷ cost
Read it for the last 30 days and the last 14, so you see the direction as well as the level.
Starting target: between the two
At or a little under achieved ROAS, and always above break-even. Move it later in small steps.
What is target ROAS in Google Ads?
Target ROAS is a Google Ads Smart Bidding strategy that sets bids to reach an average conversion value for each dollar of ad spend. Google’s help page on Target ROAS bidding defines the target as “the average conversion value (for example, revenue) you’d like to get for each dollar you spend on ads” and uses a shoe store that wants $5 of sales per $1 of spend, a 500% target, as the example.
Two details on that same page get skipped in most guides. Search and Shopping campaigns need at least 15 conversions in the past 30 days before target ROAS is available. And Google warns that “setting a target that’s too high may limit the amount of traffic your ads may get.” Most of the damage I find in ecommerce accounts comes from ignoring that one sentence.
How do I calculate a target ROAS that makes money?
A profitable target ROAS sits between two numbers you can calculate today: break-even ROAS from your margin, and the ROAS the campaign has actually achieved. Start with the floor. If gross margin after product cost, shipping and payment fees is 60%, break-even ROAS is 1 ÷ 0.60, about 167%; a target below that buys sales at a loss.
Then pull conversion value and cost for the last 30 days. That achieved ROAS is the ceiling for a starting target, and Google uses the same logic: its recommended target “is calculated based on your actual ROAS over the last few weeks.” Here are the two accounts this article is built on.
| Clothing store | Health products store | |
|---|---|---|
| Gross margin, approx. | 60% | 53% |
| Break-even ROAS | 166% | 188% |
| Achieved ROAS when trouble started | 538% over 14 days, 558% over 30 | Reported 373% over 30 days; real 144% after removing one refunded order |
| Target that choked the campaign | 600% (757% in an earlier episode) | 400% |
| Where the target went | 500% | 220%, the band it had proven in June |
What happens when target ROAS is too high?
When target ROAS sits above what a campaign can reach, Smart Bidding lowers bids until fewer auctions qualify. You see it as falling impressions with budget left over. The cleanest example I have is a clothing store’s Performance Max campaign: in late September its target was 757% and it spent only about $55 of a $100 to $110 daily budget. On September 30 the owner lowered the target to 700%, and from October 1 to 5 the campaign spent its full budget with 565 to 824 clicks a day. Same feed, same products, same budget; one number moved.
In October the same campaign showed the second way this happens. From October 14 the target sat at 600% and nobody touched it for a week, while the achievable ROAS slid underneath it. A checkout bug had been rejecting part of the card payments for weeks, so fewer clicks turned into recorded sales, and achieved ROAS dropped to 538% over 14 days. Smart Bidding pulled back.
| Period | Impressions per day | Clicks per day | Avg. CPC |
|---|---|---|---|
| October 1-11 | about 55,000 | about 590 | $0.19 |
| October 18 | 38,000 | 409 | $0.25 |
| October 20 | 13,000 | 138 | $0.40 |
| October 21 | 4,600 | 60 | $0.77 |
Search impression share fell from about 45% to under 10%. Google reported roughly 90% of eligible impressions lost to ad rank and none lost to budget. That split is the signature of a target set too high: the money is there, the bids are not.
Why did four changes in one day make it worse?
Four target changes in one day make it worse because each one resets the bidder’s adjustment before the previous one can show a result. On October 21 the clothing account got four edits in about 13 hours: the target went from 600% to 540%, then the campaign switched to Maximize Conversions with a $35 target CPA, then back to target ROAS at 500%, and the budget went from $130 to $140. The first edit landed at 1:29 in the morning.
Each edit was reasonable on its own. Google’s guidance is that after a target change “the bidder will react immediately but will need some time to hit the new target”, and it suggests giving it 1 to 2 conversion cycles. My advice to the owner had been one edit, 600% down to about 500%: inside what the campaign had proven and still three times break-even. The final setting landed in the same place. The route there cost a day of learning, and as I write this on October 22 the campaign has not recovered yet.
Does bad conversion data break target ROAS?
Yes. Target ROAS optimizes toward whatever conversion value Google Ads holds, so a wrong value trains the bidder as surely as a wrong target. In July 2026 a health products store took a $5,800 order (40 packs of eye drops) that was refunded in full two days later. The refund happened in WooCommerce. Google Ads never heard about it, because a refund does not fire the purchase tag in reverse.
That single order made one week read ROAS 9.69. The owner, reading the dashboard in good faith, raised the target to 330% and then to 400%. The campaign could not get near it, and daily spend fell from $174 to about $30 for most of August. Reported 30-day ROAS said 373%; without the refunded order it was 144%, under the 188% break-even.
The fix had two parts. We uploaded a retraction for the refunded order, which Google supports for “returned purchases” through conversion adjustments (it only works if the original conversion was sent with an order ID). Then the target went back to 220%, the band the store had actually achieved in June.
How should I change target ROAS?
Change target ROAS one small step at a time, after checking the data and the change history, and then leave it alone for one to two conversion cycles. The steps below are the order we follow on every account.
- Open Change history first. Before you diagnose a drop, list every edit from the last 30 days, mobile app included. In one account a single tap on a recommendation changed the budget from $110 to $143 and set a 140% target in the same second, and it only showed up when we pulled the full change log.
- Compare the target with achieved ROAS for 14 and 30 days. If the target is above both, lower it toward the achieved number.
- Make one change, then wait 1 to 2 conversion cycles, as Google recommends. If buyers usually purchase within three days of the click, that means about a week.
- Move gradually. Google’s own advice for more volume is to consider “gradually reducing the target ROAS”; I keep single steps to 10 to 15%.
- Do not judge one day. Google can spend up to twice the average daily budget on a single day and balances it over the month, never charging more than 30.4 days’ worth.
- Clean conversion data before you touch the target: refunds, test orders and duplicate tags all distort the value Smart Bidding learns from.
Put the numbers where everyone can see them
The calmest accounts I manage have one shared note with three lines: break-even ROAS, achieved ROAS for 30 days, current target. When everyone with access to the account looks at the same three numbers, the 1 a.m. edits stop.
If you want a second pair of eyes on yours, that is what our Google Ads management work starts with.
Campaign spending less every day?
Send us the target, the budget and 30 days of results. We will tell you whether the target, the data or the auction is the problem.