What is MER?

By the HonestTag team · Published July 12, 2026

Every advertising metric answers a question. MER answers exactly one: for every dollar the business spent on marketing, how many dollars of net revenue did the whole business produce? It does not answer which ad, which channel, or which audience. That refusal is the point. A number with no attribution model in it is a number no attribution model can inflate.

The formula

MER = total net revenue / total ad spend, over the same window.

Both sides need honest definitions, because most of the dishonesty in MER lives in the definitions.

• Revenue means net sales excluding tax: gross sales minus discounts minus returns, plus shipping collected. Refunds land on the day the refund occurs, not back on the order date. Money returned this week reduces this week's number.
• Spend means all cash out: platform spend pulled from your connected ad accounts, plus manually added costs like agency retainers and influencer flat fees. If you paid it to market the store, it belongs in the denominator.

Defined this way, MER is the closest thing marketing has to a bank statement: revenue you actually kept, divided by money you actually spent.

What MER is for, and what it is not

MER is the company scoreboard, and it moves slowly. Treat it as a quarter-grade metric, best read over 30, 60, and 90 day windows. A single hot week tells you almost nothing. A 90-day trend tells you whether the whole marketing machine is getting more efficient or less.

What MER cannot do is assign credit. It cannot tell you which channel or campaign did the work, because it is deliberately attribution-free. That is a feature, not a gap. The moment a metric starts assigning credit, it inherits every weakness of the model doing the assigning. MER stays out of that fight entirely, which is exactly why it is trustworthy as a top-line reading and useless as a budget-allocation tool. Use it for the first job. Refuse to use it for the second.

NMER: the version that runs acquisition

The most important relative of MER is NMER, new-customer MER: new-customer net revenue divided by total ad spend. Some operators know the same number as Acquisition MER, or aMER (one number, two labels); HonestTag labels it NMER.

NMER exists because blended MER contains revenue your brand would have earned anyway. Returning customers come back through email, through habit, through typing your name into a search bar. When that revenue sits in the numerator, prospecting looks better than it is. NMER strips it out and shows what paid acquisition brought in. MER reads the whole business; NMER isolates what paid acquisition brought in.

The pair to watch when scaling

Here is the trap that catches good operators: spend rises, NMER falls, and someone panics.

NMER falling under rising spend is expected. Efficiency per dollar dilutes as you push into colder audiences, while total acquisition grows. The honest read is the pair: NMER x new-customer count. If NMER slips from 2.1 to 1.8 while monthly new customers grow from 400 to 700, the machine is working. Judging a scale-up by NMER alone causes panic cuts of spend that was doing its job. Watch the product of the two, and watch first-order AOV alongside it, because a shift in what new customers spend changes what any efficiency ratio means.

MER vs ROAS

ROAS is platform-attributed revenue divided by platform spend: the platform grading its own homework. View-through credit, retargeting credit, and the platform's chosen attribution window all push the number up, and every incentive in the room points the same direction. MER-family metrics are computed from your store's back end and your bank-account reality. No attribution model can inflate them, because there is no attribution model in them. When ROAS says the ads are great and MER says the business is not growing, the two are measuring different things: ROAS is the platform's attributed view, MER is the store's actual bank reality.

The dishonest versions

MER is simple, which makes the cheats easy to spot once you know them:

• Blended MER used to justify prospecting budgets. Returning revenue flatters the number, so struggling acquisition hides behind loyal customers.
• MER on gross revenue, tax included. You do not keep the tax. Counting it inflates every window.
• Refund-blind MER. A week looks great until the returns arrive. Honest MER puts them back in, on the day the refund occurs.
• The quietly shrinking denominator. "Brand" spend, agency fees, and influencer payments slip out of the spend line, and the ratio improves without the business improving.

What each metric reads

1. MER reads company health and cash. It is the number a finance conversation runs on.
2. NMER, paired with NCAC, reads acquisition. Together they see through the returning-revenue haze.
3. A single week of MER reads as noise. Refund timing and platform spend restatements make weekly MER jumpy by nature.
4. Under rising spend, NMER x new customers reads truer than NMER alone. Diluting efficiency with growing acquisition is what scaling looks like.

In HonestTag, the acquisition metrics feed verdicts (Scale, Hold, Cut, or No verdict when the evidence is too thin), so the reading and the decision live in the same place.

How HonestTag computes it

HonestTag computes MER and NMER from your store's own Shopify order data. Every order is classified new or returning against store order history. New customers attribute to the first ad click, and every attribution stores a proof record: click id, timestamps, match method. Refunds reduce store-level revenue metrics. Google Ads and Microsoft Advertising receive the available retraction or restatement, and Klaviyo and Google Analytics 4 receive a refund event. HonestTag does not send Meta, TikTok, Pinterest, Snapchat, Reddit or OpenAI Ads a refund or retraction event; the refund is recorded in order proof. Spend is pulled from connected platforms, plus manual spend lines for everything else you pay for. Every threshold and formula is published in numbers on the in-app methodology page, and correctness is never paywalled. The same math runs on Starter as on Scale+, and no plan buys a different answer. Terms used along the way are defined in the glossary.

Frequently asked questions