Marketing Measurement Maturity Assessment
Is your measurement telling you the truth? Seven questions, two minutes, no email.
0 of 7 answered
Answer all 7 questions to see your result.
Why platform numbers drift from reality
Every ad platform is graded on the revenue it can claim, so every platform is built to claim generously. The most common way it happens is retargeting over-crediting: you serve an ad to someone who was already heading to checkout or about to fill out your form, they convert, and the platform books the conversion as if the ad created it. Much of that volume would have arrived with no ad at all, but it lands squarely in your reported ROAS or CPA.
View-through inflation compounds it. A user scrolls past an impression, never clicks, buys two weeks later for unrelated reasons, and the platform still attributes the purchase to the view. Branded search behaves the same way: it harvests demand you already generated elsewhere and reports it as fresh performance. None of this is fraud. It is the predictable result of letting the seller mark its own exam.
The practical consequence is that two advertisers with identical dashboards can be running very different businesses. One is genuinely growing, the other is paying to take credit for demand that already existed. The dashboard cannot tell them apart. Only a test can.
What mature measurement looks like
Mature teams never rely on a single source of truth. They triangulate. Platform ROAS or CPA tells you what each channel claims, blended metrics tell you what the whole account is doing, and lift tests tell you what actually changed when the ads ran. The three rarely agree, and the gaps between them are exactly where the insight lives.
Marketing efficiency ratio, or MER, is the anchor. Because it divides total revenue by total spend, it cannot be gamed by attribution windows or double-counted across channels. When MER holds steady while a platform reports a spike, you know the spike is a measurement artifact rather than real growth. For lead generation the same anchor is blended CPA: total spend divided by total new customers, regardless of which channel claims them. Reconciling platform numbers against finance or your CRM every month keeps that anchor honest.
Lift tests turn the picture from directional to decisive. A holdout or geo experiment gives you an incremental ROAS, or a true cost per incremental conversion, you can trust, and once you know the true incremental return of a channel you can reallocate budget toward marginal gains instead of toward whichever platform reports the friendliest number. That is the difference between measuring activity and measuring impact.
Frequently asked questions
What does this assessment measure?
It gauges how much you can trust your reported paid media numbers, whether you measure in ROAS, CPA, or both. The seven questions cover how you judge performance, whether you run lift tests, how much you lean on retargeting, whether platform numbers reconcile with finance or your CRM, and how you set budgets. Together they place you in one of three maturity bands, from platform-dependent to measurement-mature.
How does the scoring work?
Each answer carries a point value from zero to three, weighted so that the practices most predictive of honest measurement, such as running holdout tests and reconciling with finance, count for more. Your points add up to a score out of eighteen, and that score maps to a band with a tailored read on where your blind spots are and what to fix first.
What is incrementality testing?
Incrementality testing measures the sales a channel actually caused rather than the sales it merely reported. You hold out a randomized or geographic control group from your ads, compare it to the exposed group, and the difference is the incremental lift. It is the only way to separate spend that drove growth from spend that took credit for demand that already existed.
Do I need to give an email, and is it free?
No email, no sign-up, and no cost. You answer the questions in your browser, your responses stay private and are never stored, and your result appears instantly. The whole thing takes about two minutes.