Blended MER vs ROAS

Admaxxer is a DTC analytics platform with built-in Meta + Google ad ops. Most DTC teams treat MER and ROAS as interchangeable — they are not. The short answer: ROAS is a *platform-reported* ratio of claimed revenue to spend on a single channel, and it is almost always inflated by overlap; MER is the *blended* ratio of all revenue to all paid spend, and it's the number that actually predicts your P&L. ## TL;DR - MER = total revenue / total paid spend (across all channels, all sources) - ROAS = platform-reported revenue / platform spend (channel-specific, usually inflated) - When channels overlap (Meta retargeting + Google brand), ROAS sums above reality - MER under-credits channel-level performance; it hides which channel is dragging - Lead with MER at the P&L level, ROAS at the creative-decision level ## MER and ROAS defined precisely Marketing Efficiency Ratio (MER) is `total_revenue / total_paid_spend`. If you did $250k in revenue on $80k of paid spend across Meta + Google + TikTok, your MER is 3.13. That's unitless and platform-agnostic. You can calculate it without trusting any platform's attribution claims. Return on Ad Spend (ROAS) is `platform_reported_revenue / platform_spend`. Meta will tell you its ROAS is 4.2; Google will tell you its ROAS is 5.1; TikTok will say 3.8. These are platform-self-reported numbers and, if you sum them up with the spend-weighted average, you will get something like 4.4 — a number that will *always* be higher than your real MER if any channels overlap. The gap between blended ROAS and MER is the overlap tax. See the [Admaxxer blended MER tile](/docs/metrics/mer) for a real-time calculation that pulls revenue from Shopify webhooks and spend from each connected platform. ## When ROAS lies ROAS over-reports whenever two or more platforms claim the same conversion. The classic case: a customer sees a Meta ad on Monday, searches your brand name on Google Tuesday, and buys Wednesday. Meta will claim that purchase via its 7-day-click attribution. Google will claim it via its last-click on brand search. Both platforms report positive ROAS on the same $80 order. If Meta claims $80 and Google claims $80, their combined reported revenue is $160 — but your Shopify receipt is $80. ROAS-sum is 2×, MER shows the truth. This overlap tax scales with: - **Retargeting intensity** (Meta-heavy retargeting on Google-driven traffic inflates Meta ROAS) - **Brand search spend** (Google brand campaigns steal credit from Meta prospecting) - **Email + paid overlap** (Klaviyo emails + paid retargeting both claim the same reactivation) On mature DTC accounts, the sum of platform-reported ROAS is typically 1.4× – 1.8× the true blended MER. That gap is what you're paying for the convenience of platform-native reporting. ## When MER lies MER is a scalar over your whole business, and it has one big blind spot: it can't tell you which channel is dragging. If your MER drops from 3.2 to 2.8 over a month, MER alone won't say whether Meta CPMs jumped, Google brand CTR fell, or your conversion rate dropped. You need channel-level ROAS, or better, the Admaxxer [channel contribution MMM](/docs/mmm) to decompose MER into per-channel contribution. MER also masks profitability when your channel mix shifts. A move from 60% Meta / 40% Google to 40% Meta / 60% Google can keep MER flat while shifting you into higher-margin or lower-margin customer segments (Google brand search typically has higher LTV; Meta prospecting typically lower). If you optimise only on MER, you can accidentally shift into lower-value mix with no alarm. ## Diagnostic steps ### Step 1: Pull the overlap tax In the Admaxxer dashboard, calculate `(sum of platform ROAS × platform spend) / total revenue` — that ratio minus 1 is your overlap tax percentage. If it's over 35%, your platform-reported ROAS numbers are fiction at the sum level. ### Step 2: Benchmark against industry MER - Apparel: 3.0–4.5× is healthy, <2.5 is unprofitable - Supplements: 2.2–3.5× is healthy (higher margin masks lower ratio) - Furniture: 1.8–2.5× is healthy (long consideration, high AOV) - Beauty: 2.5–3.8× is healthy If your MER is below these bands, the problem is real regardless of what platform ROAS says. ### Step 3: Run incrementality on your biggest channel Platform ROAS is a *correlation*. Incremental ROAS (iROAS) is the *causal* version — the revenue that wouldn't have happened without the spend. Admaxxer's [incrementality module](/docs/incrementality) runs a two-proportion z-test on paid vs. organic cohorts. On brand search, iROAS is often 0.2× the reported ROAS. On prospecting, it's often 0.7–0.9×. ### Step 4: Reconcile MER to contribution margin MER of 3× doesn't mean profitable. If your contribution margin (after COGS + fulfillment + payment fees) is 30%, MER of 3.33× is breakeven. Your target MER should be `1 / contribution_margin` at minimum to avoid losing money on paid acquisition. ## Related signals Three diagnostics often correlate with MER changes: - **New customer rate** — if MER drops but new customer rate rises, you're paying to acquire; healthy - **Repeat rate** — if MER drops and repeat rate also drops, email or retention is breaking, not ads - **AOV** — if MER drops while AOV grows, you're scaling into lower-intent audiences; investigate before cutting spend These are all in the Admaxxer [cohort LTV](/docs/metrics/ltv) and [new vs repeat revenue](/docs/metrics/new-vs-repeat) tiles. The [Claude agent](/docs/ai-agent) can pull all three in a single `query_metrics` call. ## FAQs **Is blended ROAS the same as MER?** Close but not identical. "Blended ROAS" usually means the weighted sum of platform-reported ROAS, which is inflated by overlap. MER uses total revenue as the numerator — cleaner and harder to fool. **What MER target should I aim for?** It depends on your contribution margin. A rough rule: target MER = 1.3 / contribution_margin to allow for returns, chargebacks, and model error. **Does MER work for subscription businesses?** Yes, but use first-order MER plus LTV:CAC separately. MER alone undercounts the lifetime value of a subscriber. **Can the Claude agent calculate MER automatically?** Yes — `query_metrics` with `metric: "mer"` returns blended MER for any date range, and the agent will explain which channel caused the change if MER shifts more than 10%.