ROAS vs POAS vs MER: Which Metric to Steer By

A 400 percent ROAS looks like a campaign that prints money. At a 20 percent contribution margin, it still loses money on every euro you put behind it. Whether you steer your ad budget by ROAS, POAS or MER decides whether you end the month with profit or just with revenue.

ROAS tells you how much revenue a platform credits to each unit of ad spend, POAS tells you how much contribution margin that spend actually produced, and MER compares total revenue with total marketing cost. For campaign-level bidding, POAS is the more honest target as soon as your margins vary noticeably across products. You still need MER on top as a cross-channel check, because every ad platform counts conversions by its own rules.

ROAS vs POAS vs MER: definitions and formulas

All three metrics divide some kind of return by some kind of marketing cost. What separates them is which return and which cost go into the equation.

ROAS (return on ad spend) = conversion value ÷ ad spend. Google's Target ROAS help page walks through it as a percentage: $5 in sales for every $1 in ad spend gives you a 500 percent target ROAS. The conversion value is whatever your tracking reports to the platform, usually the order value. That is the first catch. An order value tells you nothing about cost of goods, shipping, payment fees or returns.

POAS (profit on ad spend) = contribution margin of the attributed conversions ÷ ad spend. A POAS of 1.0 means your ads produced exactly as much contribution margin as they cost. Anything above that pays towards overheads and profit; anything below burns cash. The key decision is which costs you deduct. The sensible set is every variable cost per order: cost of goods, shipping, payment fees and an expected returns rate.

MER (marketing efficiency ratio), often called blended ROAS = total revenue ÷ total marketing cost for a period. HubSpot frames MER as a business-level metric, while ROAS measures at channel level. The big advantage: MER does not care which channel claims which conversion.

How to calculate your break-even ROAS

Before anyone debates targets, you need a floor. The formula is simple:

Break-even ROAS = 1 ÷ contribution margin rate

Your contribution margin rate is the share of revenue left after all variable costs. Here is a worked example for an online store, with illustrative numbers only:

Net revenue of an average order: €100
Cost of goods €45, shipping and packaging €8, payment fees €2, expected cost of returns €5
Contribution margin before advertising: €40, or 40 percent

Your break-even ROAS is 1 ÷ 0.40 = 2.5, or 250 percent. Below that, every extra order you buy through ads costs more than it brings in. Make sure revenue and costs sit on the same basis. If your tag sends gross values including VAT (19 percent in Germany, different elsewhere), your platform ROAS looks better than it really is by exactly that factor. If you calculate break-even on net figures, either convert the platform ROAS to net or send net values in the first place.

For new-customer campaigns, you can deliberately set a target below break-even if you know what a customer is worth across repeat purchases. That is an investment decision, and it should rest on solid repeat-purchase data rather than gut feeling.

When POAS steers better than ROAS

ROAS treats every unit of revenue the same. That works as long as your margins are similar across the range. Once they spread out, ROAS systematically pushes budget towards the wrong products. Another illustrative example: two product groups both deliver a ROAS of 4.0. Group A earns a 50 percent contribution margin, group B only 15 percent. POAS for A is 4.0 × 0.50 = 2.0; for B it is 4.0 × 0.15 = 0.6. To your bid strategy they look identical. Economically, A is paying for B's losses.

This is where POAS earns its keep. On its online sales resource page, Google states that customers using gross profit optimisation in Smart Bidding see on average a 15 percent uplift in campaign profit compared with revenue-only optimisation. The footnote cites internal Google data from Shopping campaigns, September to October 2023, with profit defined as revenue minus cost of goods sold. That is a vendor figure, not an independent test, but the direction makes sense.

POAS pays off most when you sell many products with different margins, when discounts or bundles squeeze margin, or when shipping costs vary a lot by basket. For a SaaS product with a single price point, POAS adds little over ROAS. There, the more important question is what a lead or trial is actually worth.

MER and blended ROAS: the cross-channel view

Every ad platform grades its own homework, and each one uses its own rules. According to Google's help centre, data-driven attribution in Google Ads distributes credit across interactions with Search, Shopping, YouTube, Display and Demand Gen ads. For website conversions, Meta by default counts conversions up to 7 days after a link click, one day after an engagement such as a like or comment, and one day after a mere impression, as Jon Loomer documented in detail after Meta's change in March 2026. A customer who sees a Meta ad and then buys through a Google ad can easily show up in both reports.

The result: add up the conversion values of all platforms and you often land above your actual revenue. MER corrects for that because both numerator and denominator come from your own systems, your store or ERP revenue and the sum of all marketing costs. An illustrative calculation: €180,000 net revenue in a month on €30,000 of marketing cost gives you an MER of 6.0. If Google and Meta together report €210,000 of conversion value, you know at least €30,000 is counted twice, and part of the rest might have happened without ads anyway.

MER has a floor as well. Divide 1 by your average contribution margin rate and you get the MER at which marketing just pays for itself. Because MER also includes existing customers, organic sales and seasonality, it is no tool for steering individual campaigns. Think of it as your early-warning system: if platform ROAS goes up while MER goes down, you are probably buying conversions that would have come anyway.

The limits of platform attribution

Attribution distributes credit; it does not prove cause. Google describes data-driven attribution as comparing the paths of users who convert with those who do not. That reveals patterns, not proof that the conversion would have been lost without the ad. For that question you need incrementality testing. Google offers Conversion Lift, an experiment with a treatment and a control group that, according to its help page, is enabled through your Google account representative. For budget allocation across channels there are marketing mix models. Google opened its open-source model Meridian to everyone on 29 January 2025. It can take results from incrementality experiments as priors and run budget scenarios on profit, not just sales.

In practice: platform ROAS and POAS are the steering inputs for the algorithm, MER is the check for leadership, and incrementality tests are the referee when the two disagree. Clean data is the precondition for all of it, which is why your tracking setup comes first, not last.

How to put POAS to work in Google Ads

There is no POAS switch in Google Ads. You get there by changing the conversion value that Smart Bidding optimises for. Google documents three ways to do it:

Send contribution margin as the conversion value. In its conversion value best practices, Google explicitly lists profit margins as a value you can maximise with Maximize Conversion Value and a target ROAS, and it recommends transaction-specific values when each sale is worth something different to you. Calculate the margin per order in your store backend or server-side and pass it to the tag instead of revenue. A 100 percent target ROAS then equals break-even, and a 130 percent target means 30 cents of margin above ad cost for every unit spent.
Use conversion value rules. According to Google's help centre, they adjust conversion values for reporting and Smart Bidding based on audience, location or device. The API documentation shows they can add to, multiply or set a value. They are too coarse for product-level margins but useful for, say, valuing new-customer audiences differently from existing customers.
Conversions with cart data plus cost of goods. If you send cart data and add the cost_of_goods_sold attribute in Merchant Center, Google Ads reports gross profit per campaign. Google also describes a profit optimisation goal for Performance Max and Standard Shopping that uses this data, labelled as a beta on the Google page cited below.

When you switch from revenue to margin, the scale of your values changes overnight. An existing 400 percent target ROAS no longer fits. Google recommends recording conversion values for at least four weeks or one to two conversion cycles, whichever is longer, before setting a target ROAS. Treat the switch as its own test phase, ideally starting with a single campaign. Setting up bidding cleanly across channels afterwards is at the core of performance marketing.

For B2B lead generation the same logic applies one step earlier: instead of revenue, you pass an expected value per lead, derived from close rate, average deal size and margin. Google recommends enhanced conversions for leads and frequent uploads of offline conversions for this.

FAQ

What is a good ROAS?

There is no universally good ROAS. What matters is your break-even ROAS, which is 1 divided by your contribution margin rate. At a 25 percent margin it is 4.0; at 50 percent it is 2.0. Derive your targets from there, not from industry comparisons.

Does MER replace ROAS?

No. MER tells you whether your marketing works economically as a whole, but it cannot steer a single campaign. ROAS or POAS remain the targets for bid strategies. Put both side by side every month and explain the gaps.

Can I use POAS without a full store integration?

Roughly, yes. You can work with an average margin rate per product group or enter an estimated cost of goods in Merchant Center. Google's help page gives an estimate of 80 percent of the product price as an example. The more your margins vary, the more an exact calculation per order pays off.

Sources

Want more of the attention you get to turn into customers? Let’s spend 15 minutes on your setup, free and without a pitch deck. Book your free call

ROAS, POAS or MER: Which Metric Should Really Steer Your Ad Budget