Key takeaways
For a board-level MER, treat total marketing spend as a finance-agreed accounting policy rather than a dashboard default. The defensible default denominator includes paid media, agency fees, creator and influencer landed cost (organic fees, production, usage rights, related agency or platform fees, gifted product and logistics), creative production, and marketing software, pulled from the ledger. Keep discounts, promo codes, refund credits, COGS, fulfilment and payment fees out of that pool, they belong in contribution margin or as contra-revenue. Rule the contested lines in writing: loaded marketing headcount (my default is to exclude from headline MER and disclose separately), affiliate commissions (expense or CM3, never both), and retention-platform fees. Keep brand and performance inside one all-in denominator and show the split underneath. Pair the spend policy with a net-revenue numerator, an accrual period basis that matches retainers, prepayments and returns lag, and a change-control rule that restates comparatives. Report MER beside contribution profit, or do not report it.
The decision is not whether to put MER in the board pack. It is what you are prepared to defend as 'total marketing spend' when a CFO or NED asks why last month's ratio moved.
Two competent teams can close the same trading month and produce two different MERs. One divides commerce revenue by paid media. The other divides the same revenue by paid media plus agency fees, creator costs, production and marketing software. Vendor explainers treat that gap as a loose denominator versus a strict one. One illustration has the same brand reading 7.0 on the loose version and 4.2 on the strict version. Those figures are a worked example, not a benchmark, and they still make the operating point: an unstated inclusion list means the board is not reading efficiency.
Platform ROAS remains the in-account steering metric. MER is the number that is supposed to survive outside the ad account. That only holds if marketing and finance have already agreed the pool.
What does an unstated spend definition actually cost?
An inconsistent definition makes period comparison unreliable. If agency retainers, creator usage rights or a new attribution subscription enter the pool in March and nobody restates January and February, the ratio will look as if media efficiency collapsed when the only change was the chart of accounts.
When finance challenges a MER movement, I treat it as a definition fight first: what was coded into the denominator, not a media-efficiency story. Period comparison is unreliable until that ruling is written. I would rather have restated numbers than a clean-looking MER that changed meaning mid-year.
Misclassification cuts the other way too. Marketing expenses sit on the income statement as operating costs, typically inside SG&A, though some companies cut a dedicated marketing and advertising line so stakeholders can see the activity. That presentational choice affects how easily anyone can analyse marketing spend. If the MER denominator does not match the ledger view finance already uses, the board pack and the statutory numbers diverge, audits get harder, and a 'better MER' may be cheaper media or a quieter reclassification.
The commercial cost is a growth plan built on a ratio the business cannot reconcile. Hiring, brand investment and media budget increases all inherit the error. MER measures revenue efficiency, not profit efficiency. If the denominator is a presentation choice, the board cannot use it to judge whether paid growth is worth buying.
Why do the usual shortcuts fail?
The usual workaround is to report blended ROAS and call it MER.
They are not the same denominator. Blended ROAS, in the narrow sense most operators mean, is revenue over paid-media spend. MER, in the board sense, is revenue over a defined marketing-cost pool that can include agency fees, creative production, influencer payments and software subscriptions. Teams use the terms interchangeably because the gap only bites when non-ad costs are material, the difference between a modest overlay on media and a cost base that is a large fraction of it. Until you know which number a vendor, agency or dashboard is showing, you cannot compare it to your own. Ask what is in the numerator as well as the denominator before you import anyone else's ratio.
How to read blended media efficiency without losing the plot is the companion problem: even a clean blended media signal still has to be read beside contribution, not as a substitute for it.
The second shortcut is the ads-only MER. It is tidy, it matches platform exports, and it is only a reasonable approximation while non-ad marketing cost is immaterial. The moment retainers, creators and tools become a real line, the shortcut stops being conservative and starts flattering the ratio.
The third shortcut is excluding brand spend from the headline denominator so performance MER still 'looks efficient'. That does not measure brand. It hides an investment the business has already made.
The fourth shortcut is folding COGS, shipping and payment fees into 'total marketing spend' and calling the result a more complete MER. That conflates a marketing efficiency ratio with contribution margin. Gross margin is not contribution margin: payment fees, fulfilment, returns, discounts, sales commissions and variable support also move with the order. Those costs belong in the margin structure, not in the marketing pool. A ratio that mixes them cannot be compared with any other team's MER, and it will not tell you whether media is efficient.
The fifth shortcut is changing the definition when the number gets awkward. Whether agency fees and tools belong is not standardised. The wrong move is changing the definition often and regularly.
Is total marketing spend a formula or a policy?
Treat it as an accounting policy that happens to feed a ratio.
The formula is trivial: revenue divided by total marketing spend. The work is agreeing, in writing, which revenue streams and which marketing costs sit inside that formula, then holding the definition still long enough for a trend to mean something. Inconsistent definitions distort trends. Every function that touches the number, media, brand, finance, the agency, has to use the same definitions, or the argument in the boardroom is about classification rather than trading.
That agreement starts in the chart of accounts, not in an ads manager. Pull spend from the ledger. Platform dashboards are useful for in-flight decisions; they are a poor source of truth for a board denominator because they omit retainers, production, tools and anything that never passed through the pixel.
I co-own VIBAe. Penang Media is the agency. The VIBAe case page on penangmedia.com publishes a MER-above-3 reinvestment rule. That is the published note, not a house-wide target.
How should you define the denominator, the numerator and the period?
What belongs in the default inclusion list?
For a board pack, use the strict pool. Independent explainers converge on the same core lines even though they do not publish a standard.
Paid media. Auction spend across search, social and shopping, plus any creator-identity advertising or whitelisting that is actually spent in an ad account. Keep organic creator fees and paid delivery on separate lines so the same pound is not counted twice.
Agency fees and retainers. Media, creative and influencer-agency retainers that exist to produce or place the work in the pool.
Creator and influencer cost. The organic creator fee plus production, usage rights, exclusivity, agency or platform fees, product and logistics, measurement, legal review and contingency, priced as landed cost, not as the post fee alone. A cheaper post fee can still be a more expensive campaign once rights and delivery are included.
Creative production. Shoots, editing, UGC production paid to studios or freelancers, and usage extensions that are not already in the creator line.
Marketing software. Attribution platforms, ad-tech, and the subscriptions whose job is to plan, buy, measure or produce marketing.
That default is stricter than paid media only, and it is the version worth taking upstairs, because it is the version that still makes sense when non-ad cost is material.
How should you rule the five contested lines?
There is no industry consensus on these. Say so in the policy. Then rule them.
Loaded marketing headcount. Some strict definitions sometimes include the loaded cost of the marketing team. There is no survey evidence in the public sources for how common that is, and there is no accounting-body standard in this dossier that requires it. My own editorial position, not an industry standard, is to keep salaries and on-costs out of headline MER and disclose a fully loaded view as a separate line. The board can see both. The headline ratio remains comparable if you hire or freeze headcount. If you do load salaries, do it for the whole year and restate.
Affiliate and creator commissions. Accounting guidance aimed at US GAAP draws a useful distinction even if it is not your local framework: consideration paid to a customer and tied to the sale reduces the transaction price; consideration to a third party who is not the customer is usually an expense. Affiliate treatment is where judgement enters. Promo codes, referral credits and volume rebates tied to the transaction belong in contra-revenue, not in marketing spend. Coding a discount as a marketing expense overstates both revenue and spend. Affiliate partner commission is more often an expense, and some contribution-margin frameworks sit affiliate payouts in CM3 alongside customer-service cost and chargebacks rather than in marketing spend. Agree the treatment with your accountant, under IFRS or FRS 102 as applicable. Never count the same pound in the MER denominator and in contribution margin.
Gifted product cost and shipping. These are part of landed influencer cost. Include them in the creator line if you include creators at all, at COGS plus outbound shipping, not at retail. Do not also leave those units in product COGS if you have already moved them into marketing.
Usage rights, whitelisting and production. Price them separately from the organic fee. Whitelisted spend that hits the ad account belongs in paid media; the rights fee that made the asset usable belongs with the creator or production line. Split them so you can see whether you are paying for content, for rights, or for auction delivery.
Retention, email and SMS platform fees. If the tool exists to send marketing, it is a marketing software cost and belongs in the strict pool. If you prefer to treat lifecycle platforms as an operating cost of the customer base, that is a defensible alternative, but then the policy must say so, and the same rule must apply to every comparable period. Do not toggle the line on in a month when you want MER to look worse, or off when you want it to look better.
Those board-pack defaults are still not a signed ruling. One unsigned working preference, not a house rule, is to include affiliate commission if it paid for the sale, keep gifted-product cost in margin rather than in MER, and leave loaded headcount, usage rights and retention tooling out unless the board asked for fully loaded marketing. Until someone signs that, period comparison stays unreliable.
Useful for
- A written ruling on each contested line stops month-on-month MER movement from being a reclassification
- Separate disclosure of fully loaded headcount lets the board see efficiency without breaking the headline trend
- Keeping discounts and customer incentives out of the denominator protects both revenue and spend from being overstated
Watch for
- There is no surveyed standard for headcount, tooling or affiliate commission, so the ruling is yours to defend
- A strict pool will read as a weaker MER than the ads-only number the growth team already watches
- Affiliate commission sitting in CM3 rather than in MER is a genuine disagreement between measurement frameworks
What must not sit in total marketing spend?
Keep out of the denominator anything that is a reduction of the selling price or an order-variable cost of fulfilling demand:
- Discounts, promo codes, referral credits and volume rebates tied to the transaction
- Refunds and returns credits
- Payment processing fees
- Fulfilment, shipping subsidies and COGS
- Sales commissions and variable support
Those lines decide whether the order was worth taking. They belong in contribution profit. Putting them in MER produces a hybrid no other operator can compare against, and it hides the actual media-and-marketing cost the board asked for.
Should brand spend sit inside the headline ratio?
Yes. Split the reporting, not the definition.
Run one all-in denominator. Underneath it, show brand versus performance as a sub-split so brand investment is visible rather than quietly excluded to protect the ratio. Excluding brand spend inflates MER. Lagged payback is a reporting-design problem, rolling windows, cohort notes, a distinction between in-period and carry-over, not a reason to remove the cost. No source in this research set gives an evidenced ecommerce brand-to-performance ratio worth quoting, so do not invent one for the pack.
Creator work that is really paid amplification should not be filed as 'brand' to keep it out of performance MER, and auction spend against a creator handle should not be filed as 'influencer' if it already sits in the media invoice.
Which revenue belongs in the numerator?
A precise denominator is worthless if the revenue line wobbles.
Pick net of refunds, returns, discounts and shipping pass-through for board reporting, and write that down. Gross revenue flatters MER. Growth teams default to gross because it is close to the commerce dashboard; finance defaults to net because it is close to the P&L. That disagreement is the point of the policy.
Returns lag matters. One explainer argues that returns lag two to four weeks in most ecommerce operations, so a weekly MER built on gross orders includes revenue that has not yet come back. The same piece uses a worked illustration of a brand with a 24 per cent return rate reporting 4.6x against a possible true 2.9x once returns are recognised. Treat that as an illustration of the distortion, not as your number. Monthly or quarterly MER, with returns matched into the period, is the board version. Weekly gross MER is an operations pulse, and it overstates efficiency.
Scope the revenue streams in the same paragraph. Online DTC is the usual core; retail, wholesale and marketplace need an explicit in-or-out. If those channels do not share the marketing-cost pool, either keep them out of the numerator or allocate spend on a rule you can defend. Public sources barely go beyond naming those streams, so reason from matching: the numerator should be the revenue the denominator was spent to drive.
The preference here, not a signed ruling, is that board MER should be DTC revenue against the spend that bought it. Retail, wholesale and marketplace stay out unless that spend is allocated to them. The current mix has not been published.
How do you stop timing from inventing a trend?
Cash timing of retainers, annual tool prepayments and lagged returns will move MER even when nothing in trading changed.
Accrue retainers across the months they cover. Spread annual software prepayments. Match prepaid media to the flights it bought. Hold returns in the period of the original order once you have a stable lag assumption, or report MER on a lag that finance already uses for revenue recognition. The period basis, calendar month, trading month, or rolling 28 days, should be the same for numerator and denominator.
Run the calculation before the board-prep cycle, from the ledger, not from a screenshot of last week's ad accounts.
What goes on the one-page policy?
The artefact the board actually needs is shorter than the argument.
- Inclusion list: paid media, agency fees, creator landed cost, production, marketing software
- Exclusion list: discounts and contra-revenue, refunds, COGS, fulfilment, payment fees, and (if you follow the default) loaded salaries
- Rulings on the contested lines, in one sentence each
- Revenue basis: net of refunds, returns, discounts and shipping pass-through; named channels in scope
- Period basis: accrual, not cash; returns lag stated; monthly or quarterly for the pack
- Owner: named marketing and finance joint owners
- Change control: definition fixed for the financial year; any change restated across comparatives and minuted
Treat MER as a rolling monthly or quarterly figure, not a year-end surprise. Consistency of definition is the requirement. A prettier definition that moves twice a year is worse than a strict one that does not.
Where does the public evidence stop?
The evidence is stronger on taxonomy than on magnitudes.
Multiple independent vendor and agency explainers agree that a broad denominator includes paid media, agency fees, creator costs, creative production and marketing software, and that a narrow denominator is paid media only. They agree that the industry has not standardised the choice. They agree that platform ROAS is the in-account metric and that MER is the one that is supposed to be finance-reconciled. They agree you should define revenue streams and marketing costs in writing because inconsistent definitions make periods incomparable.
They also agree, from different directions, that MER measures revenue efficiency rather than profit efficiency, and that a defensible floor comes from contribution-margin structure rather than from a borrowed multiple. A simplified break-even is one divided by the pre-advertising contribution margin rate, subject to returns, discounts, fulfilment, payment fees and attribution assumptions. If variable costs drift, the existing MER target is no longer correct and should be re-derived from CM2. Prioritise contribution-margin pounds over a percentage, because pounds are what cover fixed cost.
What the dossier does not support is equally important. There is no survey of how many D2C brands include agency fees, headcount or tooling. There is no IFRS, FRS 102 or ICAEW extract in this research set on marketing-cost classification; the accounting sources here are vendor guides, and the contra-revenue discussion is US GAAP-framed. There is no evidenced typical movement in reported MER when a definition changes, beyond unevidenced illustrations. There is no evidenced brand-versus-performance mix. There is no evidence of how boards or investors respond to restated definitions. Do not fill those gaps with a neighbouring blog's multiple.
Where sources disagree, keep the disagreement on the page: loaded headcount is only 'sometimes' in scope; affiliate payouts may sit in marketing spend or in CM3; an ads-only shortcut is argued as acceptable when non-ad costs are immaterial. Consensus is not available. A written ruling is.
That is also why judgement still sits with the operator, not with the dashboard. Reading how AI changes D2C research without replacing judgement is the same discipline in another corner of the stack: retrieval can organise the invoices; it cannot decide the policy.
What should you lock before the next pack goes out?
Sit with finance and write the one-page policy before you refresh the MER slide.
Put two numbers on that slide: MER on the agreed definition, and contribution profit in pounds for the same period. MER on its own is only a ratio. That pairing is what belongs on the slide. It is not a claim that every board pack is already built that way.
If you are deciding whether to buy more volume, check the economics before increasing Meta ad spend against that pair, not against platform ROAS. Whether paid media should optimise for ROAS or profit is the decision MER cannot take for you: ROAS can steer delivery inside an ad platform; profit decides whether the resulting growth is worth buying.
If lifecycle tooling sits in the pool, make sure the underlying flows are worth the subscription. Which ecommerce lifecycle flow should you fix first? is the operating question that follows the reporting one.
Recast a comparative period only if the definition changed, and then recast every period you still show. That restatement is the test. Last year's MER has not been restated onto today's definition, so the months where the story would change cannot be named. Without it, any year-on-year MER movement is still a definition risk.
