Skip to content
← All posts
Measurement ·June 24, 2026 ·3 min read

The Marketing Metrics That Actually Decide Anything

Most marketing reports are full of numbers that cannot change a decision. Here are the ones that can.

Lila Galipeau
Lila Galipeau
Founder, Partnr Media

The short answer

A metric earns its place only if a change in it would change what you do next. Revenue, contribution margin, repeat rate, CAC payback and owned-channel share pass that test. Reach, impressions and follower count usually do not.

Most marketing reports are long and decide nothing. They contain reach, impressions, engagement rate, follower growth, and a chart of website sessions — numbers that go up and down without ever prompting anyone to do something differently.

There is a simple test for whether a metric belongs in a report: if this number moved 20% in either direction, would we do something different? If not, it is context at best and decoration at worst.

The five that pass the test

1. Revenue by channel, with contribution margin. Not revenue alone. A channel producing $50,000 at 10% margin is worth less than one producing $30,000 at 40%, and revenue-only reporting hides that permanently.

2. Repeat purchase rate, by cohort. This governs what you can afford to spend acquiring a customer. Measured by cohort, not across all time, or recent customers drag the number down simply for being recent.

3. Customer acquisition cost, and how long it takes to pay back. CAC alone is meaningless without knowing what a customer is worth and when you get the money. A high CAC with a two-month payback is fine. A low CAC with a fourteen-month payback can kill a business with a working profit margin.

4. Share of revenue from owned channels. Email, SMS, direct, repeat. This is the single clearest measure of whether your demand belongs to you or to a platform that can change its rules. Watch it as a trend.

5. Conversion rate, segmented. Site-wide conversion rate is an average that hides everything. Segment by device, by traffic source, by new versus returning. That is where the actionable version lives — for example, mobile converting at a third of desktop is a specific, fixable problem.

The ones that mislead

Reach and impressions. They measure delivery, not interest. Reach can be bought and does not indicate anyone cared.

Follower count. A number that only goes up, which is precisely why it feels good and tells you nothing. Engaged audience matters; total does not.

Open rate. Since privacy features started pre-loading images, a meaningful share of recorded opens are synthetic. Use it directionally for deliverability, never as a performance measure. Click and placed-order rates are real.

Engagement rate in isolation. A post can be highly engaging and sell nothing. Worth watching as a leading indicator of reach, not as an outcome.

Last-click attribution, taken literally. It reliably over-credits the final touch — branded search, retargeting, coupon sites — and under-credits everything that created the demand in the first place.

Two habits worth more than any dashboard

Look at cohorts, not totals. Total revenue tells you what happened. Cohorts tell you whether it is getting better. A business with flat revenue and improving cohorts is in a much stronger position than the reverse, and a total-only view cannot distinguish them.

Write down what you expect before you look. Stating the expectation first turns reporting into a test rather than a search for a flattering story. It is uncomfortable and it is the fastest way to stop fooling yourself.

One dashboard, read the same way

Reporting fails most often not because the numbers are wrong but because everyone has their own version. Platform dashboards disagree with each other by design — each one claims as much credit as its attribution window allows.

Pick one source of truth, define each metric in writing, and have everyone read the same thing. The definitions matter more than the tooling: “conversion rate” means nothing until you have agreed on the denominator.

What to do about the numbers that flatter you

Every business has a metric that reliably looks good, and it tends to get more airtime than it deserves.

The discipline is to lead the report with the number that decides the most, whether or not it is doing well. If revenue is flat, the report opens with flat revenue — not with a 340% increase in reach.

Where this sits in the framework

This is Results & Relationship, the sixth stage — and the reason it is a named stage rather than an afterthought is that most agencies report on what makes them look good.

Vanity metrics climb, the bank account stays flat, and nobody can point to the moment it went wrong. Reporting on the numbers that decide things is the only version of measurement that is worth the time it takes.

Let's build your trust growth system.

Trust is the asset. PARTNR is the system. Bring us the one thing capping your growth right now — we'll work it through together on the call.

Free 30-minute call. If we are not the right fit, we will tell you — and point you somewhere better.