Target ROAS on a store with 30 orders a month: start low or match past ROAS?

Say a store does around 30 orders a month and wants to move from Maximize Conversion Value to Target ROAS. Their last 90 days averaged 420% ROAS, so the obvious move is setting the target at 400% and letting Smart Bidding take it from there.

I think that often backfires. With that few conversions the algorithm has little to learn from, and the auction has probably shifted since those numbers were earned. Starting around 250 to 300% and stepping up as the conversion count builds gives it room to find volume. The catch is that starting low risks overspending on weak sales, and reports can look fine while margin quietly disappears.

So when the client is checking the bill every week, which do you pick? Curious how people handle a thin account where the target is tied to margin.

Iris, start from margin, not past ROAS. Break-even ROAS is just 1 divided by gross margin, so a store with 35% margin after COGS breaks even around 285%. A 420% average can easily hide low-margin products that lose money at a 400% target. Starting at 250 to 300% might work for that store, but it's really a guess unless it lands near break-even. Margin labels by product group in the feed will tell you a lot more than the blended average. Do they already have margin in the feed, or is that the first thing to build?

Priya, the math holds, but 1/margin only works if the margin is what the store actually keeps after the sale. Gross margin after COGS usually still has payment fees, free shipping and returns coming out of it. Say a store keeps 35% gross but loses another 5 points to fees and shipping. Net is 30%, so break-even is 1/0.30, about 333%, not 285%. That gap matters when you build the target from it.

The conversion value in the tag matters too. If it includes VAT or shipping charged to the customer, part of that 420% is money the store never keeps, so the real break-even moves again. Check what's inside that value before you compare it to any margin figure.

Theo, I'd push back on turning fees into a flat 5 points. Payment fees are mostly per order, so they hit a €40 basket much harder than a €200 one. Say the store has a €40 average order, 35% gross margin, and a payment fee of 2.9% plus €0.30. That comes to about €1.46 per order, or 3.6 points, and that's before any shipping the store absorbs on orders under the free-shipping threshold. The percentage moves with basket size, so break-even moves too. I'd run the numbers on the AOV of the segment you're bidding on, not the blended figure.

Returns are different. I'd handle them through conversion adjustments on the value rather than averaging a return rate into margin, because return rates often vary a lot by product, and that's exactly what the blended number hides.