The Right Way to Measure Marketing ROI (Without Fooling Yourself)

Marketing teams measure ROI constantly. The problem is not a lack of data — it is that the data most commonly reported does not accurately represent what marketing is actually producing.

Measurement that feels rigorous but is fundamentally flawed leads to worse decisions than honest uncertainty would. At least with honest uncertainty you know what you do not know.

The Last-Click Lie

Most marketing attribution gives full credit for a conversion to the last touchpoint before the sale. Someone reads three blog posts, sees a retargeting ad, attends a webinar, and then clicks a branded paid search ad before signing up. The paid search ad gets 100% credit. The content and retargeting get none.

Decisions get made on the basis of this attribution. Content budgets get cut. Retargeting gets scaled. Branded search spending increases. The top-of-funnel work that created the demand slowly disappears, demand drops, and the branded search ads that were “working” stop producing because there is no longer any awareness to capture.

Last-click attribution does not measure the marketing that drives growth. It measures the final step in a journey that earlier marketing made possible.

Vanity Metrics and What They Hide

Impressions, reach, follower counts, email open rates — these are all metrics that can trend favourably while the business underperforms.

This does not make them useless. It makes them insufficient as primary success measures.

An open rate of 45% on an email sequence sounds impressive. But if the emails are opened and not acted upon, the high open rate indicates a compelling subject line and a weak body, or an audience that is interested but not qualified. Optimising for open rates will make that metric better while the thing you actually care about — clicks, sign-ups, revenue — stays flat.

The test for any metric: if this improves while revenue declines, is this still a good metric? If yes, it is a leading indicator worth tracking. If no, it is a vanity metric.

The Attribution Problem for Long Sales Cycles

B2B businesses with sales cycles of three, six, or twelve months have a fundamental attribution challenge. The marketing touchpoint that created the lead happened months before the deal closed. By the time the deal is in the CRM, the connection to specific campaigns is difficult to trace.

The most practical solution is a combination of approaches:

Pipeline sourcing: Tag every lead with its original source at the point of entry. Even if conversion attribution is complicated, understanding which channels are filling the pipeline tells you where demand is being created.

Cohort analysis: Track conversion rates for leads that entered in a given month or quarter, then observe their progression over time. This surfaces how different channel cohorts perform over the full cycle, not just at a snapshot.

Self-reported attribution: Ask customers, simply and directly, how they heard about you. The data is imperfect but the signal is often cleaner than platform attribution, which is affected by ad blockers, cross-device journeys, and cookie expiry.

Revenue Is Not Always the Right Metric

This sounds counterintuitive. Revenue is the metric. But attributing revenue directly to marketing channels is complicated by sales cycle length, sales team effectiveness, pricing changes, and product quality — most of which marketing does not control.

Intermediate metrics that marketing does control — lead volume, lead quality, pipeline contribution, brand search growth, organic traffic — are often more actionable than revenue attribution in the short term, provided they are chosen because they actually predict revenue, not because they are easy to report.

The discipline is in keeping the chain of logic visible. If we improve this metric, why do we believe it will eventually improve revenue? If you cannot answer that clearly, the metric is not being measured for the right reason.

What Good Measurement Actually Looks Like

Simple, honest, connected to decisions. A small set of metrics that marketing controls, that have a credible relationship to business outcomes, reviewed consistently enough to learn from trends rather than noise.

No measurement system is perfect. The goal is to be wrong in less costly ways — to have enough signal to make better decisions than you would without any measurement at all.

---

PNK WORKS builds marketing measurement frameworks before a campaign goes live, not after. Start a project.