Chapter 13
POAS instead of ROAS: Steering advertising by profit
Table of contents
Chapter 4 showed why ROAS is misleading. This chapter goes one step further: how to not only evaluate advertising by POAS but steer it accordingly – all the way to automated bidding on contribution margin.
The three maturity levels
Level 1 – POAS in the report: POAS is calculated monthly in hindsight. Steering is still by ROAS, but the managing director sees what is really left.
Level 2 – POAS per channel and campaign: POAS is broken down by channel, campaign type and product group. Budgets are shifted specifically to the profitable campaigns.
Level 3 – Bidding on contribution margin: instead of revenue, the ad platform receives contribution margin as the conversion value and automatically optimizes its bids for it. Your margin then influences bids in the auction itself.
What POAS per campaign reveals
An electronics store with three main campaigns in Google Ads. Overall ROAS almost 9, presented as solid in the report.
Campaign | Revenue | Cost | ROAS | CM margin | POAS |
|---|---|---|---|---|---|
Brand search | €120,000 | €8,000 | 15.0 | 22% | 3.30 |
Shopping general | €180,000 | €25,000 | 7.2 | 18% | 1.30 |
Shopping accessories | €60,000 | €7,500 | 8.0 | 42% | 3.36 |
Total | €360,000 | €40,500 | 8.9 | 23.3% | 2.07 |
Two insights that ROAS does not show:
Accessories deliver the highest average POAS, even though their ROAS is lower than brand search. The high margin makes the difference. This is where to test first whether additional budget still brings extra return.
Brand search looks better than it is. Many of these buyers would have bought without an ad (see Chapter 2). Its incremental POAS is significantly lower. Whether and how much budget it deserves can only be shown by a test.
The most important idea: the next euro
All the numbers so far are averages. For budget decisions, however, a different question counts: What does the next advertising euro bring in?
As budget grows, the additional return per advertising euro typically falls. A campaign with an average POAS of 3 may only reach a POAS of 1.1 on the last euro invested.
An example: an additional €5,000 of advertising generates €6,000 of additional contribution margin before advertising in a controlled test. Marginal POAS 1.2 – after advertising, €1,000 more is left. The next €5,000 only bring in €4,500 of additional contribution margin. Your result falls by €500, even though the average POAS in the report still looks brilliant.
What follows from this:
A high average POAS does not automatically mean "more budget". It means: it is worth testing whether additional budget still stays above 1.0.
The most profitable budget is reached when marginal POAS is around 1.0. Above that, every additional euro brings in more than it costs. Below that, it costs more than it brings in.
Budget increases should be tested step by step – ideally as a controlled experiment, for example with comparison regions or the ad platform's experiment features. A simple before-and-after comparison has to account for seasonality, promotions, demand and conversion lag. Its result is initially an estimate of the extra return, not proof.
Guidance for decisions
Average POAS | Assessment | Next step |
|---|---|---|
below 1.0 | Loss zone | Pause the campaign or rebuild it fundamentally |
1.0 to 1.5 | Just positive – contributes to fixed costs but is not yet profit | Only for strategic reasons, e.g. new customer acquisition |
1.5 to 2.5 | Healthy | Keep running, optimize the structure |
above 2.5 | Strong | Test more budget step by step and watch marginal POAS |
Bidding on contribution margin: the two routes
Route A – Passing contribution margin as the conversion value. At purchase, the order's contribution margin rather than its revenue is reported to Google Ads. To do this, the cost price per product is stored in the tracking, ideally server-side via server-side tracking so that your margins are not visible in the browser. The target ROAS bid strategy then automatically steers by contribution margin.
Important here: the target value has to move too. So far your tracking reports €100 of revenue; in future, at a 30 percent contribution margin, €30 of contribution margin. With the same product mix, a previous target ROAS of 500 percent then corresponds to a target of 150 percent. If the target simply stays at 500 percent, you are suddenly demanding more than three times the profitability overnight – and the campaigns collapse.
Route B – Cart data and cost of goods at Google. Google Ads can combine submitted cart data with the cost of goods stored in the Merchant Center feed via the cost of goods sold attribute. From this, Google calculates gross profit and gross margin at product level and shows them in the account. Initially, this is additional reporting – gross profit only becomes a bidding target once it is used as the conversion value for optimization. Cost of goods is taken into account, but not shipping, fees or returns.
Custom labels in Merchant Center also help group products by margin class into separate campaigns. But they do not steer bids by profit.
Combined with clean server-side tracking, bidding on contribution margin is particularly effective for assortments with very different margins – exactly where a uniform target ROAS costs the most money.
A 90-day plan
Days 1 to 30: determine contribution margin cleanly, add monthly POAS to reporting, set the break-even ROAS.
Days 31 to 60: POAS per channel, campaign and product group. Shift budgets, test brand campaigns.
Days 61 to 90: first record contribution margin values in parallel and check them against inventory management. Only once there is a reliable data history over several conversion cycles, switch the bid strategy with an adjusted target value and test budget increases in a controlled way.
How arboro puts this kind of bidding into practice in Google Ads is explained on the page Google Ads for Shopware and Shopify. The data foundation for it is described on the page server-side tracking.
What you can check today
Do you know your POAS at monthly level?
Do you know your POAS broken down by channel, campaign and product group?
Has your brand search been tested for additional revenue in the last twelve months?
Does your tracking pass contribution margins or only revenue to Google Ads?
Is cost of goods stored in the Merchant Center feed?
Are budget increases tested step by step and measured by additional contribution margin in the store?
Frequently asked questions
Profit on ad spend: contribution margin per advertising euro. Calculated as contribution margin I divided by ad spend. Above 1.0, advertising makes money.
Contribution margin I of the orders won through this channel divided by the channel's costs. The attribution of orders remains an attribution, so also work with tests.
Yes, if contribution margin is passed as a conversion value used for optimization. Cart data and cost of goods in Merchant Center initially enable gross profit reporting. Profit only becomes the bidding target once it is used as the conversion value for optimization – with an adjusted target.
The contribution margin brought in by the last advertising euro invested. It is almost always below the average POAS. The most profitable budget is reached when it is around 1.0.
An average between about 1.5 and 2.5 is healthy in many stores. If it is much higher, it is worth testing whether additional budget is still profitable.
)