MER and platform ROAS answer different questions. Platform ROAS helps an ad system and its operator decide where to route spend inside an account. MER shows whether total business revenue is keeping pace with total paid-media spend. Neither metric proves incremental profit. A useful operating system pairs MER with new-customer CAC and contribution margin, then uses platform ROAS as a diagnostic—not as the company scoreboard.
Start by defining the two metrics
Platform ROAS is the conversion value credited to a platform divided by spend on that platform. The word credited matters. Google Ads explains that its attribution model determines how conversion credit is distributed and that the selected model also affects conversion-based bidding. In other words, platform ROAS is part measurement and part optimization input—not a neutral copy of the finance ledger.
MER, or marketing efficiency ratio, is often used in e-commerce to mean total revenue divided by total paid-media spend:
MER = total business revenue ÷ total paid-media spend
Teams do not always use the term consistently. Some include agency fees, creators, affiliates, or other marketing costs in the denominator; others include only media. Name the version in the dashboard. This guide uses paid-media MER.
The numerator needs a contract too. Shopify distinguishes gross sales, net sales, and total sales, and its total-sales calculation can include taxes, duties, shipping, and fees. For a management dashboard, select a revenue definition that matches the decision—often net sales after discounts and reversals—and keep it consistent across periods.
Why the dashboards can all be “right” and still disagree
A shopper may see a Meta ad, search the brand on Google, open an email, and buy. Each reporting system observes a different portion of that path and applies its own rules.
Google Analytics defines attribution as assigning credit to touchpoints and offers several reporting models. It also notes that modeled key-event data can be updated after the conversion is recorded. Shopify's marketing reports warn that sales attributed to marketing can differ from sales in other reports. These are expected consequences of different scopes, windows, identity signals, and models—not automatically evidence that one system is broken.
That creates two common mistakes:
Adding Meta-attributed revenue and Google-attributed revenue as if the totals were mutually exclusive.
Treating a change in platform ROAS as proof that total business economics improved.
A recent discussion among paid-media operators on Reddit surfaces the same practical objection: which number should an operator use when platforms claim overlapping revenue? That language is useful because it exposes the real job. The operator does not need one metric to replace every other metric. The operator needs a hierarchy of metrics, with a defined decision attached to each one. Reddit informs the question here; it is not the evidence for the measurement recommendations.
Build a three-level source-of-truth hierarchy
Level Primary metric Decision it supports Main limitation --- --- --- --- Business Paid-media MER and contribution dollars Is the overall acquisition system economically healthy? Blends paid, organic, repeat, promotional, and seasonal effects Customer New-customer CAC and first-order contribution Are we acquiring enough new customers at an acceptable cost? Requires reliable customer classification and cost allocation Channel Platform ROAS, CPA, conversion value, and funnel signals Where should spend move inside a channel? Depends on platform attribution and modeling rules
This hierarchy does not make platform reporting disposable. Google and Meta use conversion signals to optimize delivery. Meta's Conversions API, for example, is designed to improve the connection between business events and its optimization and measurement systems. Clean signals matter. They simply do not turn an attributed conversion into proof of incrementality.
For the surrounding architecture, see Attribution beyond Meta and the server-side tracking guide.
Set targets from unit economics, not an internet benchmark
There is no universal “good MER.” A viable target depends on gross margin, fulfillment and payment costs, returns, repeat contribution, organic demand, retail or wholesale mix, and the amount of profit the business needs to retain.
Start with the customer:
first-order contribution before acquisition = net revenue − product cost − fulfillment − payment fees − other variable order costs
Then define an allowable new-customer acquisition cost:
allowable nCAC = first-order contribution + future contribution intentionally funded − required profit buffer
Be conservative with future contribution. Use a documented cohort horizon, account for refunds and churn, and separate observed behavior from a forecast. If the business cannot tolerate the payback period, an attractive lifetime-value model will not solve the cash constraint.
MER can then serve as a blended guardrail, but it should be derived from the actual revenue mix and allowable spend—not copied from another brand. A company with high repeat revenue may show a strong MER while new-customer acquisition deteriorates. That is why nCAC sits beside it.
A worked example
Consider a hypothetical DTC brand for one reporting period:
Net revenue: $240,000
Total paid-media spend: $80,000
New customers: 800
First-order contribution before acquisition per new customer: $125
Paid-media MER is 240,000 ÷ 80,000 = 3.0.
Blended new-customer CAC is 80,000 ÷ 800 = $100.
First-order contribution after acquisition is therefore $125 − $100 = $25 per new customer before fixed overhead. If the approved target is $20, the period clears that specific guardrail.
Now imagine the next period shows $250,000 in net revenue, $100,000 in spend, and 850 new customers. MER falls to 2.5 and nCAC rises to about $118—even if one platform reports a higher ROAS than before. That does not identify the cause, but it gives the team a reason to investigate before celebrating the platform result.
The investigation might uncover a change in attribution, more branded demand, a promotion, a product-mix shift, weaker site conversion, delayed refunds, or genuinely lower marginal efficiency. The blended metrics reveal the economic change; they do not diagnose it alone.
Use the disagreement as a diagnostic
Platform ROAS MER / nCAC What to do next --- --- --- Improving Improving or stable Consider a controlled scale step; watch marginal contribution and lag Improving Worsening Check attribution windows, branded or retargeting mix, repeat revenue, promotions, and conversion lag Worsening Improving Do not cut automatically; inspect under-crediting, channel assists, spend mix, and business changes Worsening Worsening Diagnose offer, creative, traffic quality, inventory, pricing, site conversion, and measurement health
This is a triage matrix, not a causal model. If a large budget decision depends on knowing whether advertising created additional demand, use a suitable experiment. The geo-holdout guide explains where that method fits and why uncertainty belongs in the result.
Derive a working MER range from the P&L
A single target can hide the tradeoff between efficiency and growth. Build a range with three boundaries instead:
Break-even boundary. The maximum media spend the business can absorb after product cost, fulfillment, payment fees, refunds, and other variable costs.
Operating target. The efficiency level that funds fixed costs and the approved contribution goal.
Growth exception. A temporary lower-efficiency range that leadership accepts in exchange for a defined acquisition or market-entry objective.
Begin with a period-level contribution model. Use the same revenue definition as the MER dashboard and subtract the costs that move with orders. Then decide how much contribution the business must retain before media spend.
For a simplified hypothetical period:
Input Amount --- ---: Net sales $300,000 Product cost $105,000 Fulfillment and payment costs $36,000 Refund and variable-support reserve $9,000 Required contribution after media $50,000
The amount available for paid media is $300,000 − $105,000 − $36,000 − $9,000 − $50,000 = $100,000. On these assumptions, the operating MER target is 300,000 ÷ 100,000 = 3.0.
This is not a recommendation that the brand should target 3.0. Change the cost structure, required contribution, revenue mix, or refund reserve and the answer changes. The point is that the target now has an auditable origin. Finance can challenge an input instead of debating a benchmark copied from a conference slide.
Next, model the target under downside conditions. What happens if conversion rate falls, product mix shifts toward lower-margin items, or refunds arrive after the reporting period? A target that only works in the base case is not a safe scaling boundary.
Finally, connect the period-level range to nCAC. If MER clears its target while nCAC misses, repeat or organic revenue may be carrying the blended result. If nCAC clears while MER misses, the business may be acquiring customers efficiently while existing-customer revenue, pricing, or another channel weakens. Both cases require diagnosis before a budget move.
Adapt the framework to the business stage
A young DTC brand with little repeat revenue
For a young brand, MER and nCAC may move closely because most revenue comes from recent acquisition. The risk is premature confidence: a short period can be dominated by launch demand, creator exposure, or a promotion. Use first-order contribution and cash payback as the primary constraints. Treat projected LTV as upside until cohorts have had enough time to mature.
The budget review should ask whether new-customer volume increased without pushing nCAC beyond the approved range, whether contribution dollars grew, and whether the site converted the additional traffic. Platform ROAS helps identify where the account changed, but the business result determines whether the scale step stays.
An established DTC brand with meaningful repeat revenue
An established brand can show a healthy MER even while acquisition weakens because prior customers continue to purchase through email, direct traffic, subscriptions, or habit. Split new-customer revenue and returning-customer revenue. Track nCAC beside total MER, and review cohort contribution separately from current-period revenue.
This does not mean repeat revenue should be removed from every management metric. It means leadership should be able to see whether current media is acquiring the next cohort or whether the dashboard is harvesting value created in earlier periods.
An omnichannel brand with retail or wholesale revenue
Total company revenue divided by digital media spend can become misleading when retail distribution changes, wholesale orders are lumpy, or paid media influences stores that are not measured at the customer level. Define a digital MER for the owned e-commerce operation and a broader business efficiency view for planning. Keep the two labels explicit.
When offline impact is material, matched-market experiments, retailer data, or an appropriately designed marketing-mix model may add information that a commerce dashboard cannot. Do not solve missing scope by quietly adding unrelated revenue to the numerator.
Separate branded demand and retargeting from growth
Branded search and retargeting are useful, but their reported efficiency is easy to misread. They often reach people who already know the brand or are already close to purchase. That can make them look more efficient than prospecting without answering how much additional demand they created.
Use three checks:
Spend-mix check: Did branded or retargeting spend grow faster than prospecting spend?
Blended check: Did total revenue, contribution, and nCAC improve when that mix changed?
Experiment check: For a material allocation decision, can a holdout, suppression test, or geo design estimate what happened without the spend?
Do not impose a universal branded-search or retargeting percentage. Brand strength, competitor bidding, query mix, promotion cadence, and customer journey all matter. The control is transparency: report how much of the platform result came from existing demand and decide what role that campaign is meant to play.
Run a weekly budget review that finance can reproduce
1. Freeze metric definitions
Document the revenue field, spend accounts, currency treatment, customer definition, refund handling, reporting timezone, and data-lag policy. A dashboard that changes definitions silently cannot support a budget decision.
2. Reconcile before interpreting
Confirm that spend is complete, order totals tie to the chosen commerce or finance report, and tracking has not broken. Review the current period with enough lag for the business's conversion and return cycle.
3. Separate average from marginal performance
Average MER describes the entire spend base. The budget question is usually marginal: what happened as spend changed? Adjacent-period comparisons are useful signals, but promotions, seasonality, inventory, pricing, and channel mix can move simultaneously. Do not label the difference “incremental” without an experimental or defensible causal design.
4. Diagnose by layer
Business layer: revenue, contribution dollars, MER, cash and payback.
Customer layer: new customers, nCAC, first-order contribution, cohort retention.
Channel layer: platform ROAS, CPA, conversion value, reach, frequency, creative and funnel behavior.
5. Make bounded changes
Record the budget move, expected outcome, decision window, and stop condition. Large simultaneous changes destroy the comparison the team needs next week.
A dashboard template that prevents metric drift
The weekly view should fit on one screen. More detail can live below it, but leadership needs a stable decision layer.
Metric Current period Prior comparable period Approved range Owner --- ---: ---: ---: --- Net revenue Finance / commerce Contribution dollars after media Finance Paid-media spend Growth Paid-media MER Growth + finance New customers Analytics / commerce Blended nCAC Growth First-order contribution after acquisition Finance Platform ROAS by channel Diagnostic only Channel owner
Add annotations for promotions, price changes, inventory constraints, site releases, tracking incidents, large creative launches, and budget moves. Without those notes, the team will repeatedly rediscover the same explanation.
Use comparable time windows. A seven-day platform view beside a monthly finance total is not reconciliation. Neither is comparing order-date revenue with payment-date cash. The dashboard contract should specify the calendar, timezone, currency conversion, late-arriving conversions, refunds, and the date on which a period is considered stable enough for review.
Questions operators should ask before changing spend
Should platform ROAS be ignored?
No. It is useful for campaign diagnostics and is an input to platform bidding. The mistake is promoting it to the company-wide source of truth. Use it to understand delivery inside the platform, then reconcile the decision against business and customer economics.
Is MER better than attribution software?
It solves a different problem. MER is a blended ratio that can be reproduced from business revenue and media spend. Attribution tools allocate credit across touchpoints. A team may need both, but neither replaces contribution accounting or a causal experiment.
How often should MER be reviewed?
Use a cadence appropriate to conversion lag, purchase frequency, refunds, and spend volume. A weekly operating review is practical for many e-commerce teams, but daily movement can be too noisy and a monthly view can be too slow for active media management. Document when a period is mature enough to compare.
Can MER determine which channel caused growth?
No. It shows that blended efficiency changed; it does not isolate the cause. Channel diagnostics, controlled changes, experiments, and—where justified—marketing-mix modeling are the next layers.
What if contribution data is unavailable?
Start with a transparent approximation and label every excluded cost. Net revenue and paid-media spend are better than an unreconciled platform total, but they are not profit. Assign an owner and a deadline for adding product cost, fulfillment, payment fees, refunds, and other variable costs. Do not let the temporary proxy become the permanent definition.
When MER is the wrong headline metric
MER becomes less informative when revenue and spend do not share a sensible time horizon, when a business mixes materially different models, or when non-media costs dominate the decision.
Examples include long sales cycles, app subscriptions with delayed value, marketplaces with incomplete revenue visibility, retail-heavy CPG, and professional services where qualified pipeline matters more than immediate revenue. Adapt the numerator and customer outcome to the business. Do not force an e-commerce ratio onto a different economic system.
Even in DTC, MER is not profit, incrementality, or a channel attribution model. It is a fast, auditable pressure gauge. Its value comes from keeping platform reporting in context.
The practical next step
Build one weekly view with four rows: net revenue, contribution dollars, paid-media MER, and blended nCAC. Place platform ROAS below those rows, not above them. Assign an owner to every definition and annotate promotions, inventory constraints, pricing changes, and major budget moves.
If the numbers cannot be reconciled—or if the team is still debating whose dashboard is “right”—the measurement system needs a contract before it needs another attribution tool. Sharply Labs' growth practice connects paid acquisition, conversion signals, blended economics, and experimentation into one operating cadence.
Sources
About attribution models — Google Ads Help
Get started with attribution — Google Analytics Help
About modeled key events — Google Analytics Help
Sales reports — Shopify Help Center