Your Attribution Is Lying. Measure the Blend.
Ask Meta, Google, and your email platform how many sales each drove last month, then add up their answers. The total will comfortably exceed your actual revenue — sometimes by double. Every platform grades its own homework, every attribution model has a thumb on the scale, and the meetings where teams argue about whose dashboard is right consume more energy than the campaigns themselves. There’s a way out, and it’s older and simpler than the tooling that created the problem.
Why attribution flatters the wrong things
Last-click attribution — still the default lens in most companies — hands credit to whoever touched the customer closest to purchase. That systematically rewards the bottom of the funnel: branded search, retargeting, the cart-abandonment email. These channels look phenomenal precisely because they harvest demand that something else created. Meanwhile the things that created the demand — the podcast, the organic content, the event, the ad that introduced you six weeks ago — report terrible numbers, because attribution can’t see across devices, browsers, dinner-table conversations, or time.
Follow last-click logic faithfully and you’ll keep shifting budget toward harvesting and away from planting. It works beautifully until the field is empty — the retargeting audiences shrink, branded search dries up, and nobody can say why, because every dashboard was green the whole way down.
The ratio that can’t be gamed
Marketing Efficiency Ratio is total revenue divided by total marketing spend. All revenue, all spend — media, tools, agencies, the retainers everyone forgets. No model, no cookie, no platform’s opinion. If you spent $200K last month and revenue was $1M, your MER is 5.0. That number can’t be inflated by attribution games, because it doesn’t ask who deserves credit. It asks the only question ownership actually needs answered: for every dollar the whole marketing system consumed, what came back?
Tracked weekly against a target, MER does something channel dashboards can’t — it tells you when the system is degrading, even while every individual channel reports success. And it reframes budget conversations: instead of “Meta says its ROAS is 8,” the question becomes “we added $50K of spend and MER fell from 5.0 to 4.2 — was that trade worth it?” Sometimes it is; growth usually costs efficiency. But now it’s a decision, not an accident discovered two quarters later.
Use attribution for direction, MER for truth
None of this makes platform metrics worthless. Channel-level numbers are fine for comparing two ads, two audiences, two landing pages inside the same channel — relative signals, where the platform’s bias applies equally to both options. The failure mode is using them as absolute truth across channels, which is exactly the decision they’re least qualified to inform.
So run two layers. The team optimizes inside channels with the platform’s own numbers. Ownership reviews one blended scorecard — MER, pipeline, revenue against targets — on a weekly cadence. When a channel’s claimed performance rises but the blend doesn’t move, you’ve learned that channel is taking credit, not creating value. That’s not a measurement failure. That’s the measurement finally working.
Baron Belalov is a fractional CMO working with growth-stage and established companies globally.