What is a good ROAS for Google Shopping?

The formula is one divided by contribution margin. Contribution margin is what is left of an order after the costs that scale with it: cost of goods, the shipping you absorb, payment processing, pick and pack, and the share of orders you expect to refund or return.

Work it from an average order rather than from a single product, because Shopping traffic buys across the catalogue. Take your real average order value, subtract each of those costs, and divide what remains by the order value. That percentage is your contribution margin, and its reciprocal is the ROAS at which advertising exactly pays for itself.

Two things trip people up. The first is using gross margin, which ignores shipping and fees and produces a break-even that is too flattering. The second is forgetting returns. A category with a high return rate, footwear and apparel in particular, needs the expected return cost taken out before the margin is calculated, otherwise the campaign looks profitable in the ads account and is not profitable in the bank.

Once you have the number, everything else is a decision about how much profit you want per order versus how much volume you want. Break-even is the floor, not the goal.

Decide the gap deliberately. Break-even ROAS covers the variable costs of the order and nothing else. Rent, salaries, software, the photography for next season and the profit you actually want to keep all have to come out of the margin above break-even.

A reasonable way to set it: take your monthly fixed costs, decide what share of them paid acquisition should carry, and work backwards to the ROAS that produces that contribution at your expected spend. A store where paid traffic is one channel among several can run closer to break-even because organic and email revenue carry the overheads. A store where paid is the only source of new customers has to run further above it.

The other input is customer lifetime value. If a first order at break-even reliably produces a second and third order with no acquisition cost attached, a Shopping campaign that just washes its face on the first purchase is doing useful work. That only holds if you can show the repeat behaviour in your own order data, cohort by cohort, rather than assuming it.

Write the target down with the reasoning attached. Targets that drift because a month looked bad are the main reason accounts stall.

Raising the ROAS target raises efficiency and lowers profit past a certain point, because Smart Bidding responds by buying less traffic. The auctions it drops first are the marginal ones, which are also where a good deal of the incremental revenue lives.

Think in total contribution rather than ratio. A campaign spending 10,000 at a ROAS of 3 returns 30,000 in revenue. At a contribution margin of 35 percent, that is 10,500 of contribution against 10,000 of spend. Push the target to 5 and spend might fall to 4,000 at 20,000 of revenue: 7,000 of contribution against 4,000 of spend. The ratio looks far better and the business kept less money.

So test the target in both directions and watch contribution, not ROAS alone. Change it in steps rather than in jumps, give each step enough conversions to be readable, and expect a few weeks of movement each time the target changes.

The exception is a store with a cash constraint or a supply constraint. If you cannot fund the working capital or ship the volume, a high target is the correct answer even though it leaves contribution on the table.

A blended Shopping ROAS hides the two things it is made of. Queries carrying your brand name convert at a much higher rate because the shopper already decided; queries describing a product category convert at a lower rate because the shopper is still choosing. Averaging them produces a number that describes neither.

Separate them so the targets can differ. In Performance Max that means brand exclusions and the account level negative list, plus a separate campaign or search themes structure for brand demand. In Standard Shopping it means campaign priorities with negative keywords, or a dedicated brand campaign.

Then set a lower target on the non-brand campaign than on the brand one, because non-brand is where new customers come from and it will never match the efficiency of people typing your name. If the two are pooled, Smart Bidding meets the blended target by leaning on brand traffic and the store slowly stops acquiring anyone new.

The cross-check for all of this is marketing efficiency ratio: total revenue divided by total marketing spend across every channel. If channel ROAS improves while the store's overall efficiency ratio does not, the campaign has been taking credit rather than creating demand.