POASvsMERvsROAS:WhichMetricforWhichDecision

POAS, MER and ROAS measure different things. Used interchangeably — which most ecommerce agencies do — they generate contradictory recommendations and undermine commercial decision-making. This article defines each metric cleanly, maps each one to the decision it should drive, and shows why mixing them up creates the kind of "we hit our ROAS target but our profit went down" outcome that has become depressingly common in 2026.

If you've ever sat in a meeting where the marketing team and the finance team were looking at different dashboards and reaching opposite conclusions about the same campaign, this article is for you.

The three metrics, defined

ROAS — Return on Ad Spend. Revenue divided by ad spend, calculated at the platform level (Google Ads, Meta, TikTok). The default success metric Google Ads surfaces. Measures the gross sales generated for every pound spent on advertising. Excludes cost of goods, fees, returns, discounts, fulfilment — everything below the revenue line.

POAS — Profit on Ad Spend. Contribution margin divided by ad spend, calculated at the campaign or SKU level. Contribution margin = revenue minus COGS, shipping, payment fees, discounts, return costs. Measures actual profit generated per ad pound, not revenue. POAS is a registered trademark of ProfitMetrics; the term is used here as the industry-standard shorthand.

MER — Marketing Efficiency Ratio. Total revenue divided by total marketing spend (all channels combined), calculated at the business level. Measures how efficiently the entire marketing function generates revenue, regardless of attribution or channel. Born out of the post-iOS 14 attribution mess as a "blended" alternative to channel-level ROAS.

The first thing to notice: each metric calculates at a different level of granularity. ROAS at the campaign level. POAS at the campaign or SKU level. MER at the business level. This is why they cannot be used interchangeably — they answer different questions.

What each metric is actually measuring

The most useful way to think about these metrics is as diagnostic tools at three different altitudes.

ROAS measures the platform. It tells you whether a specific Google Ads or Meta campaign is generating gross revenue at a rate the platform considers efficient. It's a useful diagnostic for platform-level decisions — is this campaign delivering enough revenue per click for the bid strategy to remain stable? Should this asset group be paused or scaled? It is not, and has never been, a measure of whether the campaign is contributing to your bottom line.

POAS measures the campaign's commercial outcome. It tells you whether the campaign is generating actual contribution margin after all the costs that ROAS ignores. POAS is the right metric for bid-level decisions — should we bid up on this SKU? Is this campaign profitable enough to scale? It's the metric that maps cleanly to whether the ad spend is moving the business forward financially.

MER measures the marketing function. It tells you whether your total marketing investment is generating enough total revenue to be efficient, blending across channels and removing the noise of attribution. MER is the right metric for the leadership conversation about marketing as a function — is the brand spending too much or too little overall, given current revenue? It says nothing about which campaign is working and which is not.

Each metric is useful at its altitude. Each is misleading when used at the wrong altitude.

The decision matrix — which metric for which question

This is the part most teams get wrong. The metric you pick depends entirely on what decision you're trying to make.

The pattern: ROAS for platform-level diagnostics, POAS for campaign-level commercial decisions, MER for business-level marketing efficiency. Each metric for the question it was built to answer.

Why mixing them up creates bad decisions

A real example, anonymised. A skincare brand running £30k/month on Google Ads. The reporting agency tracked ROAS at the campaign level (showing healthy 5.2x average) and MER at the business level (showing 4.1x — strong by category standards). Both numbers looked good.

POAS at the campaign level wasn't being tracked. When we audited the account, we discovered that three of the seven campaigns had POAS below 1:1 — they were losing money on every ad pound — but were being kept alive because their ROAS contribution propped up the blended average and their revenue contribution propped up the MER.

The agency wasn't lying. They were optimising honestly against the metrics they had agreed to report on. But because the metrics they had agreed to report on were the wrong granularity for the decision being made (which campaigns to scale and which to cut), the optimisation produced a perverse result: the unprofitable campaigns grew while the profitable ones were starved of budget.

This is the structural problem with metric mismatch. It's not that ROAS or MER are wrong metrics. It's that using them to answer questions they weren't built to answer produces decisions that look defensible in a deck and indefensible in a P&L.

The right reporting stack for an ecommerce CFO

If you're a founder, CFO or Head of Performance trying to set up reporting that actually supports good commercial decisions, here's what should be on the dashboard:

Top level (monthly, business-wide):

• MER (trailing 30 days, trailing 90 days, year-over-year)
• Total marketing spend by channel
• Contribution margin from marketing-attributed revenue (where attribution is possible)

Middle level (weekly, channel-specific):

• POAS by Google Ads campaign
• POAS by Meta campaign
• POAS by other channels where margin data can be attributed
• Branded vs non-branded POAS within Google Ads

Bottom level (daily/weekly, campaign/SKU-specific):

• ROAS by campaign as a directional indicator
• POAS by SKU for the top 20–50 SKUs by spend
• New customer POAS vs returning customer POAS (where LTV data is integrated)
• Feed health by SKU (disapproval rate, impression share)

The dashboard should make it impossible to look at MER without also seeing POAS, and impossible to look at POAS without also seeing ROAS as a diagnostic underneath. Each metric stays visible at its right altitude. Decision-makers stop being able to make decisions using the wrong altitude.

When MER actually helps (and when it's a smokescreen)

MER became popular in 2021–2022 because iOS 14 broke much of the attribution that channel-level metrics had depended on. The argument was: stop chasing attribution accuracy, blend everything, measure efficiency at the business level. Sensible response to a real problem.

But MER also became a way for agencies and in-house teams to obscure channel-level performance problems. If your Google Ads is delivering 1.2:1 POAS (losing money) but your MER is 4.1:1 (because organic and email are propping it up), MER is technically telling the truth while obscuring a real problem. The Google Ads spend is bleeding, but it doesn't show in the headline number.

MER is the right metric for the question "are we over-spending on marketing as a function?" It is the wrong metric for the question "is our paid search team doing a good job?"

If you only have one metric in your monthly leadership update, MER is a worse choice than POAS. POAS tells you whether the marketing function is generating profit. MER tells you whether revenue is keeping pace with spend — which can stay healthy for a long time even as profit erodes.

What to do this week

Three actions, in order of leverage:

1. Audit your current dashboard. What metric does your agency report on? If it's ROAS-only, you're getting platform-level diagnostics dressed up as commercial reporting. Ask for POAS by campaign in next month's report. If they can't produce it within 30 days, you have a structural information problem to solve.
2. Confirm your COGS data is in your ecommerce platform. POAS calculation only works if cost-of-goods is populated per variant in Shopify (or your equivalent). Most stores have 5–15% of variants with missing or wrong COGS. Audit it. Fix it. This is the foundation of every other metric improvement.
3. Pick the right altitude for the next decision in front of you. If you're deciding whether to scale a specific campaign — POAS. If you're deciding whether to shift budget between channels — MER (with attribution caveats). If you're diagnosing whether bidding is technically functioning — ROAS as a low-altitude tool. Each metric for its right job.

The brands that win in UK ecommerce in 2026 won't be the ones with the highest ROAS, or even the ones with the highest POAS. They'll be the ones who use each metric for the decision it's built to support, and never confuse one for the other.

Frequently asked questions

Next steps

Want help building a reporting stack that surfaces the right metric at the right altitude? We audit ecommerce Google Ads accounts spending £10k+/month and show you exactly where the metric mismatch is costing you.