Skip to main content Stan Consulting LLC · Marketing Atlas · Position · The Limits of MER as a Performance Metric

Stan Consulting · Marketing Atlas · Position · MER

The Limits of MER as a Performance Metric.

Marketing Efficiency Ratio is the right boardroom metric and the wrong channel-level metric. Treating MER as the sole performance metric over-rewards branded baseline and under-credits top-of-funnel investment. The position names the four-metric framework that keeps MER honest at the boardroom and channel layers simultaneously.

Read the Position → Read the Case File
01 Section 01 · The claim The claim.

MER is the right operator-side metric for boardroom defense and the wrong metric for channel-level decisions. Treating it as the sole performance metric over-rewards branded baseline acquisition and under-credits top-of-funnel investment. The fix is reading MER alongside three companion metrics, with MER as the constraint and the companions as the channel-decision drivers.

The claim has two parts. The first is mathematical: MER is total revenue divided by total marketing spend. The numerator includes every order regardless of source; the denominator includes every dollar regardless of channel. The ratio compresses information about which channel produced which dollar into a single blended quantity. At the boardroom layer, the compression is useful. At the channel layer, the compression erases the information the channel decision needs.

The second part is operational: an operator who optimizes for MER alone discovers that the highest-MER channels are usually branded search and email. Both surfaces harvest demand the rest of the marketing program created. Increasing investment in those surfaces lifts MER on the dashboard and degrades the demand-creation pipeline that fed them. The operator next quarter has a higher MER and a smaller pool of demand to harvest. The quarter after that, MER drops and the operator cannot defend why.

The position is not "ignore MER." MER stays as the boardroom number. The position is read MER as a constraint, not as a target. Read three companion metrics for the channel decisions MER was never built to inform.

02 Section 02 · The conventional view What most people believe.

The conventional read on MER is that it is the unbiased blended metric. The argument runs roughly as follows: platform-level attribution is broken since iOS-fourteen and the cookie deprecation, GA4 disagrees with the platforms, the platforms disagree with each other, and the operator needs a single defensible number that ignores the disagreement. MER does that. It is computed against revenue and spend the operator already owns, and it is immune to the attribution arguments at the channel level. The metric became the default boardroom number across DTC during the post-iOS-fourteen period for exactly that reason.

Belief 01

"MER is the unbiased blended metric." The argument is that MER reads against owned data on both sides of the ratio and avoids the platform-attribution disputes. True at the aggregate. False at the channel. MER is unbiased about which channel produced the dollar in the sense that it does not even ask the question. The blended-but-unbiased framing reads as rigor and is the source of most operator over-reliance on the metric.

Belief 02

"MER reconciles platform-attribution disputes by ignoring them." The argument is that the operator can stop arguing about whether Google or Meta gets credit and just read MER. The dispute is reconciled. Sort of. The dispute is not reconciled; the dispute is bypassed. The operator who reads MER alone does not know which channel to scale, throttle, or test. The bypass reads as pragmatism and is actually a deferral.

Belief 03

"MER is the metric to optimize against." The argument is that since MER is the boardroom metric, the team should be optimizing toward it. This is the most expensive belief in the set. Optimizing toward MER on its own pushes spend into branded baseline, email, and SMS, where MER is highest. The pipeline that supplies the baseline shrinks. Two quarters later the baseline starts to drop. The operator concludes the funnel is failing. The funnel is fine; the demand-creation layer was harvested into.

Belief 04

"MER targets are the right way to scope marketing budget." The argument is that the CFO sets a target MER and the marketing team is funded to hit it. Plausible-sounding. Operationally weak. The MER target is a constraint the marketing team has to operate inside. It is not a budget framework. Scoping marketing budget against an MER target without companion metrics produces a budget that can hit the target by under-investing in the future. The hit-the-target budget and the build-the-business budget can diverge significantly.

Each belief is supported by a real practice and a real precedent. None of them, on their own, are a defensible reason to read MER as the channel-level metric. The position is the four-metric frame that puts MER back in its place.

03 Section 03 · Why the conventional view fails Why that belief fails.

The structural argument is that MER is true at the aggregate and lies at the channel. The aggregate truth is that revenue divided by spend is the operator's blended efficiency. The channel-level lie is that this number, on its own, contains no information about which channel deserves the next budget dollar.

Five failure modes follow.

Failure mode one. MER over-rewards branded inheritance. A channel that converts demand the operator's other channels created (branded search, email, SMS, push) shows a high MER because the conversions cost almost nothing per dollar. The operator increasing branded-search spend lifts MER and reduces the demand-creation budget that built the brand the customer was searching for. The channel-level read is missing entirely from the MER number. Without companion metrics, the operator cannot tell branded harvesting apart from demand creation.

Failure mode two. MER under-credits top-of-funnel. Top-of-funnel paid social, paid video, podcast sponsorship, and influencer programs build future demand. They do not produce immediate conversions at a high enough rate to lift MER on a single quarter's read. An operator optimizing toward MER under-funds them. Two quarters later the branded baseline drops, the email list grows more slowly, and the operator who hit MER target last quarter has to justify why the baseline is shrinking. The justification is the absent investment.

Failure mode three. MER is sensitive to mix shifts that look like efficiency changes. A holiday season lifts MER because branded inheritance compounds. A January slump drops MER because demand creation has not refilled the pipeline. The operator who reads MER as a quality signal will conclude the funnel is improving in November and degrading in January when neither is actually true. The mix shifted; MER reflected the mix; the operator misread the cause.

Failure mode four. MER targets produce hit-the-number behavior. When MER becomes the headline number the marketing director is scored against, the path of least resistance is to cut top-of-funnel spend until MER hits the target. The cut works for a quarter or two. The damage compounds in the cohort layer the operator did not measure. By the time the damage is visible at the cohort layer, the marketing director who hit the MER target has been promoted and the new marketing director inherits a baseline that is shrinking.

Failure mode five. MER hides incrementality collapse. A channel can lift MER and produce zero incremental revenue. Branded search on a brand the customer was already searching for is the canonical example. The MER read says the channel is efficient. The incrementality read says the channel is harvesting purchases the customer would have made anyway. Without an incrementality read alongside MER, the operator over-funds harvesting and under-funds creation. The two reads disagree on every account that has run an incrementality test on its branded surfaces.

The conventional view treats MER as the metric. The structural reality is that MER is one of four metrics the operator needs to read together. The position is the framework that names the other three.

04 Section 04 · The SC position The SC position.

Read MER alongside three companions. MER is the constraint. Incrementality, new-vs-returning mix, and cohort LTV are the channel-decision drivers. MER stays as the boardroom number. The companions stay at the channel and quarterly layer. The discipline is the four-metric read in concert.

Each metric is named below with its scope, its source, its role, and the test that says it has been read correctly.

M1

MER as boardroom constraint

MER is the boardroom efficiency number. Defined as Shopify-net (or operator-source-of-truth) revenue divided by total marketing spend in the same window. Read as a constraint, not as a target. The constraint defines the floor MER cannot drop below before triggering an investigation. It does not define the channel mix the operator runs to stay above it.

  • Source · source-of-truth revenue / total marketing spend in same window
  • Cadence · monthly board read
  • Role · constraint floor, not optimization target
  • Companion metric required · cannot be read alone for channel decisions

Test it has been read correctly: the operator can answer "what would happen to MER if we moved this channel by ten percent" without using MER as the only input.

M2

Incrementality lift per channel

Incrementality is the lift a channel produces above the no-spend counterfactual. Geo-holdout, ghost-bid, or platform-native incrementality test. Each channel gets its own lift coefficient. The coefficient is the channel-decision input MER does not provide. A channel with high MER and zero incremental lift is harvesting; a channel with lower MER and meaningful incremental lift is creating.

  • Source · geo-holdout test, ghost-bid test, or platform-native incrementality
  • Cadence · quarterly per channel, monthly for new channels under test
  • Role · channel-decision input on harvesting versus creating
  • Threshold · channels below 30% incremental lift are presumed harvesters

Test it has been read correctly: the operator has a documented incremental-lift coefficient per major channel updated within the quarter.

M3

New-vs-returning customer mix

The share of orders from first-time buyers against returning buyers, per channel. The mix surfaces whether a channel is acquiring new customers or harvesting returning ones. A channel optimizing for MER alone tends to harvest. The mix metric names the harvesting versus acquiring distinction and is the input the operator runs the customer-acquisition budget against.

  • Source · Shopify customer table joined to acquisition channel
  • Cadence · monthly read at channel level
  • Role · new-customer acquisition tracking against demand creation
  • Floor · new-customer share floor per channel, set by category

Test it has been read correctly: the operator has a documented new-customer share per channel against the floor and a reallocation rule when the share collapses.

M4

Cohort LTV by acquisition source

Lifetime revenue per customer, segmented by acquisition channel. Cohort customers by acquisition month and read LTV at fixed intervals: three-month, six-month, twelve-month. The channel-level LTV read is the input that turns CAC into a defensible ceiling. Without LTV-by-source, CAC is a number with no ceiling and an operator's CAC argument is just an argument.

  • Source · Shopify customer-revenue history joined to acquisition channel
  • Cadence · quarterly read with rolling cohort comparison
  • Role · CAC ceiling input, channel-quality differentiator
  • Comparison · LTV against the prior cohort for the same channel

Test it has been read correctly: the operator has a documented LTV-by-source for the trailing twelve cohorts and a CAC ceiling per channel derived from it.

05 Section 05 · The mechanism The mechanism.

The working spec runs five blocks. MER as constraint. Incrementality. New-versus-returning mix. Cohort LTV. The disagreement protocol that resolves the four-metric reading when the metrics disagree. The whole framework installs in roughly four to six weeks on an operating Shopify Plus account; the reading cadence runs from there.

M1 MER as constraint Define first · boardroom layer

Compute MER and document its definition

MER is total revenue divided by total marketing spend in the same window. Document which revenue line (Shopify-net, Shopify-plus-marketplace, bank-deposited) and which spend definition (paid-only, paid-plus-agency, paid-plus-tooling) the operator runs against. The definition is part of the metric. Operators reading MER without this documentation produce numbers that drift across reporting cadences.

Set MER as the boardroom constraint

MER is the constraint number reported to the board and the CFO. Constraint means the floor MER cannot drop below before triggering an investigation, not the target the team is optimizing toward. Setting MER as a target produces under-investment in demand creation. Setting MER as a constraint preserves the CFO-side defense without distorting the channel-side decisions.

M2 Incrementality lift Define second · channel-decision layer

Compute channel-level incrementality lift

Incrementality is the lift a channel produces above the no-spend counterfactual. Geo-holdout, ghost-bid, or platform-native incrementality test. Each channel gets its own lift coefficient. The coefficient is the channel-decision input MER cannot provide. Branded search on most established brands tests below thirty percent incremental and is therefore presumptively a harvesting surface; top-of-funnel paid social typically tests above sixty percent and is presumptively a creating surface.

Document the testing cadence

Incrementality runs on a rolling cadence per channel. Quarterly for established channels, monthly for new channels under evaluation. The cadence is owned, scheduled, and documented before the test results are ever read. The cost of not having a documented cadence is that incrementality results get cherry-picked when the channel team needs them and ignored when they do not. The cadence is the discipline.

M3 New-vs-returning mix Define third · acquisition layer

Compute new versus returning customer mix

Mix is the share of orders from first-time buyers against returning buyers, per channel. The mix surfaces whether a channel is acquiring new customers or harvesting returning ones. A channel optimizing for MER alone tends to harvest. The mix read names the harvesting versus acquiring distinction and is the input the operator runs the customer-acquisition budget against.

Set a new-customer floor per channel

Each channel has a documented new-customer floor below which the channel is presumed to be over-harvesting. Brand-search budget below the floor is reallocated to top-of-funnel demand creation. The floor is set by category, not by site-wide aggregates. A subscription brand's branded floor is not the same as a one-time-purchase brand's branded floor; a fashion brand's prospecting floor is not the same as a home-goods brand's.

M4 Cohort LTV by source Define fourth · valuation layer

Compute cohort LTV by acquisition source

LTV is the lifetime revenue per acquired customer, segmented by acquisition channel. Cohort the customers by acquisition month and read LTV at fixed intervals (three-month, six-month, twelve-month). The channel-level LTV is the input that turns CAC into a meaningful constraint. The trailing twelve cohorts give the operator enough signal to read against; below twelve cohorts the variance is high enough that the read is directional.

Set CAC ceilings per channel against LTV

Each channel has a CAC ceiling derived from its cohort LTV. The CAC-to-LTV ratio is the input the operator runs the channel-level CAC budget against. Without LTV-by-source, CAC is a floating number with no defensible ceiling, and the channel-level CAC argument is just an argument. With LTV-by-source, CAC has a ceiling per channel and the channel-level decisions are scopable.

M5 Disagreement protocol Define fifth · reconciliation layer

Establish the disagreement protocol

When MER and the companions disagree, the protocol routes the conflict to a named decision. MER stable but new-customer mix collapsing: the operator is over-harvesting; reallocate to top-of-funnel. MER lifting while incrementality drops: the channel is harvesting more efficiently rather than creating more demand; freeze the channel's budget and rerun the incrementality test. MER below floor while LTV-by-source improving: hold spend at the current level and let the LTV improvement work its way through the cohort. The protocol is the document that resolves the four-metric reading on a quarterly cadence.

06 Section 06 · Evidence and case links Evidence and case links.

The Position page is the doctrine. The links below are where the doctrine has been applied or referenced for a different audience. Each link is a test the doctrine has had to pass.

Primary case

The Quarter Google, Meta, and GA4 All Claimed the Same Sale

The composite case file where the platforms each produced an inflated channel-level read. MER on this account read clean at the boardroom and contradictory at the channel layer. The four-metric framework would have surfaced the harvesting-versus-creating split the platform reads obscured.

Read the case file →

Companion case

The CFO Who Asked Why the Numbers Did Not Match the Bank

The composite case file where MER computed against the marketing dashboard read different from MER computed against Shopify-net read different from MER computed against bank-deposited. The MER definition itself was the missing artifact. Three numerators, three MER reads, one CFO question.

Read the case file →

Companion position

Attribution Is a Judgment Problem Before It Is a Tracking Problem

The companion doctrine on the three-layer attribution stack. MER sits at the boardroom layer of that frame; the channel-level reads sit at the conventional layer. The two positions read together define the firm's stance on revenue, attribution, and channel reads.

Read the position →

Adjacent position

Conversion Rate Is a Symptom, Not a Diagnosis

The adjacent doctrine on conversion rate as a four-input function. The shape of the argument repeats: a single aggregate metric (CR there, MER here) compresses information that channel-level decisions need. The decomposition is the position.

Read the position →
07 Section 07 · Where it breaks Where it breaks.

Every methodology has assumptions. Naming the assumptions is part of defending the position. The four-metric framework assumes the operator has at least six months of cohort data, a working incrementality test program, and channel-level UTM hygiene. The methodology does not handle every operator-side configuration.

01

Pre-Series-A brands without cohort data

Brands with fewer than six months of cohort data cannot read LTV-by-source against a stable cohort baseline. The methodology defaults to a rolling MER constraint plus a directional new-versus-returning read until the cohort layer becomes statistically meaningful. The full four-metric frame waits for the cohort data to mature.

02

Brands without an incrementality test program

Operators who have never run a geo-holdout, ghost-bid, or platform-native incrementality test cannot populate the M2 metric. The methodology defaults to assigning presumptive harvesting-versus-creating labels per channel based on category benchmarks until the first incrementality results land. The presumption is documented as a working hypothesis, not a measured coefficient.

03

Single-channel businesses

Operators running paid against a single channel (often a single platform Meta-only or a single-source affiliate program) cannot read mix shifts inside the four-metric frame because the mix has only one input. The methodology defaults to a CAC-LTV-and-incrementality read at the channel level without the cross-channel mix decisions.

04

One-time-purchase categories with no recurring revenue

The cohort LTV input is structurally different for one-time-purchase categories (mattresses, large appliances) than for repeat-purchase categories. The methodology defaults to a referral-and-review-driven LTV proxy or to a multi-year purchase-window read instead of the standard three-six-twelve-month cohort. The frame still applies; the LTV measurement is reshaped to the category.

08 Section 08 · What it costs to apply What it costs to apply.

The four-metric framework installs as the Conversion Second Opinion for operators who want the read on its own. The methodology is the same in either format. The deliverable shape and the engagement length are different.

Diagnostic only

Conversion Second Opinion

$99972-hour verdict

A written diagnostic verdict against the four metrics. MER read documented. Companion-metric coverage assessed. Disagreement protocol sketched against the operator's last quarter of data. No restructure, no implementation. The read.

See the engagement →

Diagnostic plus install

Sprint or System Build

Engagement-scopedread first, scope second

The diagnostic runs first as the scoping artifact. The Sprint or System Build engagement runs the install of the missing companions and the supporting reporting. Pricing is set against the install scope after the read; the read is the input that makes the price honest.

See the engagement formats →

Five Cents · Stan's note

Five Cents

The pattern I keep seeing is that MER is the CFO's favorite metric for a reason: it is computed against owned data, it is hard to argue with at the aggregate, and it closes a quarter-end conversation in two minutes. That is also exactly why it becomes a marketing trap. The CFO loves it; the marketing director adopts it as the headline; the marketing director then optimizes against it; the optimization erodes the demand-creation pipeline that supplied the high-MER surfaces in the first place. Two quarters later the CFO asks why MER is dropping and the marketing director cannot defend the answer because the answer is "we hit the target last quarter by harvesting future demand." That answer does not survive a board read.

What I want operators to take from this is that the problem is not MER. MER is fine. The problem is treating MER as the sole performance metric. The fix is the four-metric reading, where MER is the constraint and the other three drive the channel decisions. The fix is also the disagreement protocol, where the operator names in writing what to do when MER and the companions point in different directions. Without the protocol, the team defaults to whichever metric favors the action it already wanted to take. With the protocol, the team's actions become defensible.

What this position is for: if your team reports MER to the board and uses MER to set channel budgets, and you have not built the three companion reads alongside it, you have this position. The Conversion Second Opinion runs the four-metric read in seventy-two hours and produces the verdict the CFO and the marketing director can both sign off against.

Stan Tscherenkow · Marketing Atlas · 2026-05-07
10 Section 10 · Related Atlas entries Related Atlas entries.

The Reference pages in the Attribution cluster, the case files this position was written against, the companion positions, and the hub. The graph below is the cluster map.

Reference · Attribution

MER and CAC Rule

The reference page on MER and CAC. How the metrics are defined, where they break, and the rule for reading them alongside other reads. The reference foundation for the position on this page.

Read →

Reference · Attribution

Incrementality

The reference page on incrementality testing. The methodology that powers the M2 metric on this position. Geo-holdout, ghost-bid, and platform-native test design.

Read →

Reference · Attribution

GA4 Attribution

The reference page on the GA4 attribution model. The session-based stitching and last-non-direct-click default that frames how channel-level reads inform the four-metric framework.

Read →

Reference · Attribution

Data-Driven Attribution

The reference page on data-driven attribution as a methodology. Where it sits relative to MER and the companion metrics, and the conditions under which it adds signal versus noise.

Read →

Reference · Attribution

UTM Loss

The reference page on UTM loss. Where UTMs disappear, why the unattributed share grows over time, and how the loss compounds against the channel-level reads inside the four-metric framework.

Read →

Case File

The Quarter Google, Meta, and GA4 All Claimed the Same Sale

Composite case file where the platforms produced an inflated channel-level read. MER read clean at the boardroom and contradictory at the channel layer.

Read →

Case File

The CFO Who Asked Why the Numbers Did Not Match the Bank

Composite case file where MER itself read different against three different revenue numerators. The MER definition was the missing artifact.

Read →

Position

Attribution Is a Judgment Problem Before It Is a Tracking Problem

The companion doctrine. The three-layer attribution stack. MER sits at the boardroom layer of that frame; the channel-level reads sit at the conventional layer.

Read →

Position

Conversion Rate Is a Symptom, Not a Diagnosis

The adjacent doctrine. CR there, MER here, the same shape: a single aggregate metric compresses information channel-level decisions need. The decomposition is the position.

Read →

Hub

Marketing Atlas

The authority graph hub. Reference, Case Files, Positions, Indexes. The map of how this Position connects to the rest.

Read →

If your boardroom is reading MER alone

Read the four metrics together. Then steer the channels.

The Conversion Second Opinion runs this position against your account in seventy-two hours. A written verdict against the four metrics, the missing companions named, the disagreement protocol sketched against your last quarter of data. If the verdict says install, the engagement formats are scoped against the read. If the verdict says hold, you keep the read and act on it yourself.

Get the Conversion Second Opinion · $999 Or read the companion position →