MER vs ROAS: which metric should your ecommerce business use?

MER and ROAS answer different questions

Return on ad spend (ROAS) divides ad-attributed conversion value by the ad spend in scope. A Google Ads account, for example, defines ROAS as conversion value divided by cost; its conversion value depends on which actions and values you configured. A 5x ROAS means the report assigned $5 of value for each $1 of that ad spend, not that the business kept $5. Google Ads ROAS glossary · Google Ads conversion values

Marketing efficiency ratio (MER), as used here, is total store net merchandise revenue divided by all marketing cost during the same period. It includes revenue from paid, organic, direct, and returning customers without assigning orders to channels. We include media, agency, creative, and email tooling in the cost. The ratio answers: how much net merchandise revenue did the store record for each dollar of marketing cost? This revenue-over-spend convention appears in HubSpot's MER guide.

There is no universal MER label. Triple Whale's guide labels ad spend divided by order revenue as MER and presents it as a percentage. That is the inverse direction, and its spend boundary may differ. If you exchange numbers with an agency, ask for the formula rather than accepting the acronym. Our inverse is called marketing cost share: full marketing cost divided by net merchandise revenue.

Write down the denominator before comparing reports

Cost scope is where reasonable teams talk past each other. Platform ROAS usually uses spend inside one ad platform. Media-only blended ROAS often uses total store revenue divided by ad spend across platforms. That wider numerator avoids choosing an attribution model, but the denominator still excludes non-media marketing cost. Our full-cost MER adds the creative, people or agency, and tools you decide belong to marketing.

A ratio is only comparable over time if the boundaries stay fixed. Specify currency, date range, time zone, order statuses, tax and shipping treatment, refund timing, and whether salaries, affiliates, production, discounts, and software are included. Discounts reduce the revenue numerator in our example; they are not also counted as marketing cost. Record a cost once. A service business or retailer with a different cost structure can use a different boundary, provided everyone reading the chart can see it.

One illustrative month, three answers

Suppose an ecommerce store books $120,000 in discounted merchandise revenue in a calendar month and processes $10,000 of merchandise returns or reversals in that same month. For this illustrative example only, net merchandise revenue is $110,000. Tax and shipping are outside the revenue figure. The store spends $12,000 on Meta media and $8,000 on Google media. It also pays $3,000 for creative production, $2,000 in agency fees, and $1,000 for email tooling. These are invented teaching figures, not Admoji customer results or a market benchmark.

Both 5.50x and 4.23x can be arithmetically right. The difference is the $6,000 of marketing work outside media buying. If a team silently switches from media-only spend to full marketing cost, the trend appears to deteriorate even though nothing operational changed. The MER and ROAS worksheet carries these labeled inputs and a blank section for your own definitions.

Why platform ROAS does not add up to store sales

Now suppose, still illustratively, Meta reports $60,000 in attributed conversion value and Google reports $55,000. Meta's reported ROAS is $60,000 ÷ $12,000 = 5.00x; Google's is $55,000 ÷ $8,000 = 6.88x after rounding. The two reports claim $115,000 combined, while the order ledger shows $110,000 net merchandise revenue. Do not sum their credited values and call the result unique sales.

In this constructed scenario, a shopper has eligible interactions in both accounts before one purchase. If each system independently assigns credit under its own rules, the same order can appear in both reports. This illustrates a possible overlap; it does not measure any particular account. They may also use different conversion dates or assigned values, and a later return might already be in the commerce ledger but not reflected in the claim you exported. The $5,000 difference in this illustration is not an estimate of overlap; without order-level matching you cannot apportion it. Our ecommerce attribution guide shows how to reconcile actual orders before comparing credit rules. Shopify also documents how marketing reports apply attribution models to their channel views. Shopify marketing reports

Keep a platform ROAS when you are adjusting campaigns within that account, with the conversion action, attribution setting, and window attached. For a board or cash planning view, start from the store's order ledger and its full cost boundary. A platform can report useful directional performance without owning the business total.

A good ratio can still miss the margin floor

Revenue is not profit. Add an illustrative $70,000 of product and other variable costs matched to the $110,000 net merchandise revenue. The store has $40,000 contribution before marketing, a 36.36% contribution margin. Subtract the $26,000 full marketing cost and $14,000 remains before fixed overhead and taxes. That is a positive contribution in this constructed month; it is not a claim that the business is profitable.

At that assumed margin, the break-even full-cost MER for marketing contribution is 1 ÷ 36.36% = 2.75x. Equivalently, $110,000 revenue ÷ $40,000 pre-marketing contribution = 2.75. Below that ratio, marketing cost would exceed the contribution available to pay it, if the same margin held. The actual threshold changes with product mix, returns, fulfillment, payment fees, and any costs left outside the calculation.

For a marginal budget choice, use incremental economics rather than treating the average month as a forecast. At the same assumed margin, another $1,000 in marketing would need $2,750 of incremental net merchandise revenue to cover itself before fixed costs. Whether an extra $1,000 can produce that revenue is a separate question. A high ROAS on a low-margin category may be less attractive than a lower ROAS on a strong-margin category. Document the margin basis beside any break-even line so the target cannot drift away from actual unit economics.

Keep returns and time windows on the same basis

Our example places the return deduction in the same month as the booked sale to make the arithmetic readable. Real reports can treat those events on different dates. Shopify's Sales reports distinguish gross sales, discounts, sales reversals, net sales, and total sales; its reversals can be recorded on the date processed. Shopify's built-in net sales field may therefore differ from a custom order-cohort calculation or the specifically defined net merchandise revenue used here. Shopify Sales report definitions

Choose a policy before calculating MER. For a calendar-month operating view, use sales and reversals posted in that month and spending incurred in that month. For a campaign or acquisition cohort, assign later returns to the original orders and allow the cohort enough time to mature. Those views answer different questions; do not splice the revenue from one to the spend from the other. Repeat the same policy next month.

Watch other timing mismatches: an ad platform may attribute a conversion back to an interaction date, while the store records an order on checkout date. Costs can arrive as invoices after the ad ran. Record both the period represented and the export date, especially for an unsettled recent week. If a ratio moves sharply, first ask whether a batch of refunds, a late fee, or a date-setting change explains it.

Separate new and repeat customers

Whole-store MER includes every order in scope. That is useful for seeing whether total marketing cost fits the total business, but it can flatter an acquisition program when a large share of sales comes from long-standing customers. A returning buyer may purchase through email or direct traffic after earlier brand work, with little relation to this month's paid prospecting. Conversely, a new customer acquired this month may buy again later, beyond the current MER window.

Keep a simple companion view: net merchandise revenue from first orders, net merchandise revenue from repeat orders, new customers, and full marketing cost, with an explicit rule for identifying customers across orders. Where acquisition cost is the question, assign only the costs that genuinely target acquisition and document that allocation. Do not call all store revenue “new customer revenue” because a platform attributes it to a prospecting campaign. Our order management process guide covers the order status trail needed to keep these segments reliable.

Changing customer mix can move MER without a change in media effectiveness. A retention-heavy month may improve the store ratio; a deliberate first-order investment may lower it. That is a reason to read the ratio beside customer mix and contribution, not a reason to abandon it.

Use both metrics in a weekly decision routine

1. Freeze the ledger. Export eligible orders, discounts, returns, and reversals for a named period and time zone. Set the tax, shipping, currency, and cancellation rules. Reconcile the total before looking at channel claims.
2. Build the cost ledger. Add media by platform and the non-media marketing costs in your policy. Check invoices for timing and classify each line once. Calculate full-cost MER and media-only blended ROAS from the same revenue number.
3. Read platform ROAS in context. Inspect conversion action, attribution window, reported value, and spend scope. Use it to compare campaigns within a consistent setup, not as a replacement for store revenue.
4. Apply the financial filter. Review returns, contribution margin, and the first-order versus repeat-order mix. Compare the full-cost MER with your own margin-derived floor, not an unsourced industry target.
5. Test the next move. If you need to know whether increased spending caused more sales, plan an experiment or other credible incremental measurement. Google describes Conversion Lift as comparing test and control groups to estimate additional conversions or value. Google Ads Conversion Lift

Keep a dated note of definition changes. If the same ratio rises after a new creative fee was removed from the denominator, you have a reporting change, not proven improvement. Admoji's public Performance & Reporting overview shows the reporting context; use your own reconciled revenue, cost, and margin definitions when applying this routine.

Questions that settle the choice

Should I replace ROAS with MER?

No. Keep full-cost MER for the whole-business trend and the financial check. Keep platform ROAS for decisions inside a platform, where its attribution rule and conversion value are understood. Put the formulas beside the figures so the team can tell what moved.

Is MER the same as blended ROAS?

Sometimes people use the labels interchangeably. In this guide, both use total store revenue, but media-only blended ROAS divides by paid media spend while MER divides by full marketing cost. A vendor may also invert MER and report spend as a percentage of revenue. Check both the direction and the cost scope.

What is a good MER?

There is no honest universal number. Build a floor from your own contribution margin and the costs your MER denominator includes. Then inspect whether new customers, returns, and fixed overhead make the month viable. The 2.75x floor above belongs only to the illustrative 36.36% pre-marketing margin, before fixed overhead.

Do either of these prove ads created the sale?

No. MER is a period-level relationship; ROAS is attributed value per ad dollar under reporting rules. Neither supplies the missing counterfactual: sales that would have occurred without the ads. A suitable holdout or lift study addresses that question more directly. The attribution walkthrough explains that distinction using individual orders.