Profit-based bidding in Google Ads: what it needs and when it pays
The idea starts from contribution margin, the money left per order once variable costs are paid. If you do not know it per product, start there.
What does it do?
A normal Google Ads setup tells the platform how much revenue each order brought. Smart Bidding, Google's automated bidding, then aims at revenue.
Revenue is not profit. A product with a thin margin and a high price looks as good as a fat-margin one. A returned order still looks like a win unless someone tells the platform otherwise.
Profit-based bidding changes what you report back. Two pieces matter:
- Cart data and costs. Google's Search Ads 360 help describes conversions with cart data: it combines the revenue from sold items with their cost of goods sold to report gross profit2. That page covers Search Ads 360 reporting, not Google Ads bidding, so treat it as a description of the idea.
- Conversion adjustments. Google Ads Help explains that you can adjust conversions after the event. A retraction removes a conversion; a restatement changes its value3. The page sets a shorter window for adjustments to count for automated bidding than for reporting3.
The aim is POAS (profit on ad spend) in place of plain ROAS. Sources define "profit" differently, so say which definition you use.
Why do e-commerce brands use it?
Because the catalogue is uneven. Some products make a lot per order. Others barely cover their costs. Returns differ by product too.
Example (illustrative): two products both sell for €100. One leaves €50 of margin per order and the other leaves €10. Revenue bidding treats them the same. Profit bidding should not.
The same logic applies to returns. A basic ROAS figure ignores them. A return-adjusted ROAS shows what is left once they are counted. Our article on how to adjust Google Ads conversions for returns covers the reporting side.
How do we implement it?
Define profit
Pick one definition, such as contribution margin before marketing, and write it down. Gross profit and profit after fees are different numbers.
Check the cost data
Cost of goods should sit in a system and returns be tracked per order. If not, work out contribution margin per product first.
Send cart data and returns
Pass product-level data with each conversion, and a way to tell Google about returns inside the adjustment windows.
Test and maintain
Compare against what you had before. Someone keeps costs correct when prices change.
When should you not use it?
Skip it, for now, when:
In those cases, work out contribution margin per product first. That is useful without any platform feature.
What does it cost in effort?
No figures here. The work is mostly data work:
- Cost per product. Reliable cost of goods per SKU, kept up to date.
- Cart data. Product-level data sent with each conversion.
- Returns process. A way to tell Google about returns after the fact.
- Ownership. Someone who keeps the feed of costs correct when prices change.
That agency blog lists cart data and cost of goods as the prerequisites1. In our view the hard part is not the feature. It is the data behind it.
Pitfalls
Bad cost data
If costs are wrong, the bidding optimises the wrong thing at scale.
Defining profit loosely
Gross profit, contribution margin and profit after fees are different numbers. Our convention is contribution margin before marketing: revenue minus cost of goods, shipping, payment fees, returns and discounts. Pick one definition and write it down.
Ignoring timing
Returns arrive after the sale. Adjustments have time windows3.
Reading too much into the first weeks
Platform numbers still need a break-even ROAS to compare against.
Switching on without a test
Compare against what you had before.
Our take: Selective
We rate profit-based bidding Selective. This is our view, not a market fact.
Why not Core? Google documents conversion adjustments3. The profit reporting we can point to is described for Search Ads 3602, and the requirements for profit optimisation in Smart Bidding come from an agency blog, not from a Google source1. It also needs data that many brands do not have in a clean form.
Our view: when you have trustworthy cost and returns data per product, it is worth testing. When you do not, you are feeding the machine a guess.
FAQ
What is profit-based bidding in Google Ads?
It means giving Google profit information, not only revenue, so bidding can follow margin. An agency blog lists conversions with cart data and cost of goods as prerequisites1.
What are conversions with cart data in Google?
Conversions that carry the items sold in each order. Google's Search Ads 360 help says that, with cost of goods supplied, gross profit is revenue minus cost of goods2. That page is about Search Ads 360.
Can I tell Google about returns?
Yes. Google Ads Help describes adjusting conversions after the event, as retractions and restatements3.
What COGS data does profit-based bidding need?
Cost of goods per product, kept current, plus a reliable record of returns. Without both, start with contribution margin per product.
Is this the same as bidding to a target ROAS?
No. Target ROAS aims at revenue per unit of spend. Profit-based bidding tries to aim at margin.
Sources
Let’s find what’s actually working.
Next step. Want a view on whether your data is ready for this? Talk to us about your budget decision. Or read about our Paid media for ecommerce service. You can start with a free audit.