How Target ROAS works
Instead of naming a price you will pay for a conversion, you name the ratio of value to spend you want back. Google then predicts the value of each potential conversion, not just its likelihood, and bids up where the expected order is large and down where it is small. Google expresses the target as a percentage in the interface, though it is the same idea as a return multiple.
This only functions if real values reach the account. Purchases have to send a value with them, whether from a Shopify or WooCommerce integration, the data layer or the server. Send the same fixed value for every order and the strategy has nothing to distinguish between them, at which point it behaves like a cost-per-conversion target wearing a different name.
Why Target ROAS matters
For a shop with a wide spread of order sizes, treating every sale as equal wastes money at both ends: it overpays for small orders and underbids on the large ones. Value-based bidding lets the account chase the baskets worth chasing, which is why it suits catalogues where average order value varies a lot between products.
It also forces a useful conversation about what value means. Revenue is the easiest figure to send and the least honest one, because it ignores cost of goods, delivery, payment fees and returns. A ratio that looks healthy on revenue can be losing money on a low-margin product, and the account will happily keep buying more of exactly that.
Where Target ROAS goes wrong
Missing or inconsistent values are the first problem: tax and shipping included in some orders and not others, a value sent twice, or a currency mismatch will all mislead the model. Thin data is the second, because value prediction needs more history than simple conversion counting does, and a small catalogue with occasional sales rarely provides it.
Setting the target from ambition rather than evidence is the third. Ask for a ratio the account has never produced and the campaign simply stops competing. The fourth is judging success on the ratio alone: a higher ratio with far fewer sales can leave you with less money than before, which is why contribution margin is the better scoreboard. Returns compound this, since a refunded order stays in the reported value unless you send the refund back.
How to use it well
Get values right before you set a target. Decide deliberately whether the value you send is revenue or something closer to profit, apply that consistently, and check a sample of orders against your own records rather than trusting the tag. Where margins differ sharply between product groups, separate them so each can carry its own target instead of averaging good and bad together.
Then move the target gradually and read the outcome in money. Ask what total profit the campaign produced this month against last, not only what the ratio says, and expect volume and ratio to pull against each other every time you adjust it. This is the everyday discipline of running Google Ads for an online shop.