Marketing Analytics Foundations: Measure What Matters

What marketing analytics is for (in real work)

Marketing analytics is the practice of using data to make better marketing decisions: where to spend, what to change, what to stop, and what to test next. In day-to-day work, it helps you answer questions like “Is this campaign working?”, “Which channel is bringing the right customers?”, and “What should we do next week with the budget?”

A useful way to think about marketing analytics is that it supports decisions, not dashboards. A dashboard is only valuable if it changes an action (increase spend, pause a creative, adjust targeting, improve onboarding, change pricing, etc.).

Performance measurement vs. storytelling with numbers

Performance measurement is about tracking whether marketing activities are producing the intended outcomes, using metrics tied to goals. It answers: “Did we achieve what we set out to do, and what should we change?”

Storytelling with numbers is about using metrics to create a narrative that sounds good but may not change decisions. It often focuses on impressive-looking counts (views, likes, clicks) without connecting them to business outcomes.

• Measurement: “Search spend increased 20%, incremental trials increased 8%, CAC rose 5%, payback stayed under 2 months—keep spend but tighten keyword match types.”
• Storytelling: “Traffic is up 30% and impressions doubled—Search is doing great.”

The rest of this chapter is structured around three outcomes that keep analytics practical and decision-oriented.

Continue in our app.

You can listen to the audiobook with the screen off, receive a free certificate for this course, and also have access to 5,000 other free online courses.

Or continue reading below...

Download the app

Outcome 1: Identify the business question before looking at data

Most analytics mistakes start with opening a report and hunting for something interesting. Instead, start with a decision you need to make. A good business question has three parts: the decision, the options, and the success definition.

A step-by-step way to frame the question

1. Name the decision (what will you do differently?).
2. List the options (increase, decrease, reallocate, keep constant, test).
3. Define success (what outcome changes, by how much, and by when?).
4. Set the unit of analysis (campaign, channel, audience, geo, product line, cohort).
5. Choose the smallest set of metrics that can answer the question.

Examples: common marketing questions and the metrics they require

Notice that each question maps to a small set of metrics. If you need 25 metrics to answer one question, the question is probably unclear or you are mixing multiple decisions together.

Turn vague questions into decision questions

• Vague: “How are we doing on social?”
• Decision question: “Should we keep social spend flat next month, or reallocate 15% to Search?”
• Metrics: cost per incremental conversion (or CPA with strong attribution caveats), activation rate from social traffic, frequency/reach (to diagnose saturation), creative-level conversion rate.

Outcome 2: Map goals to measurable outcomes

Goals are often written as activities (“run campaigns”) or broad intentions (“increase awareness”). Analytics becomes useful when goals are translated into measurable outcomes that reflect business value.

Goal mapping: from business objective to marketing metric

Use a simple chain: Business objective → Marketing outcome → Measurable metric → Target.

Practical steps to create measurable outcomes

1. Write the goal in one sentence without marketing jargon (e.g., “Get more customers who renew”).
2. Choose the outcome closest to value (renewals, purchases, activated users), not the earliest signal (clicks).
3. Pick one primary metric (the “north” for this goal) and 2–3 supporting metrics that explain movement.
4. Define the time window (same-day conversion vs. 30-day retention).
5. Define what counts (what is a conversion, what is an activated user, what is a qualified lead).

Example: “Should we increase spend on Search?” mapped to outcomes

Business objective: Grow revenue efficiently.

Marketing outcome: Acquire incremental customers at sustainable cost.

Primary metric: Cost per incremental customer (ideal) or CPA (fallback).

Supporting metrics: Conversion rate (diagnostic), marginal ROAS (budget decision), payback period (profitability check).

How it guides action: If CPA is stable but conversion rate drops while spend rises, you may be buying lower-intent queries; action could be tightening keywords, improving ad-to-landing relevance, or capping bids on low-performing segments.

Example: “Which channel brings customers who stay?” mapped to outcomes

Business objective: Improve customer quality.

Marketing outcome: Increase retention and LTV.

Primary metric: Cohort LTV by acquisition channel (or 60/90-day retention if LTV is not available yet).

Supporting metrics: CAC by channel, payback period, churn rate by channel.

How it guides action: If Channel A has higher CAC but much higher retention and faster payback, it may deserve more budget even if it looks worse on last-click CPA.

Outcome 3: Distinguish actionable metrics from vanity metrics

An actionable metric is one you can influence with a specific lever and that reliably connects to a business outcome. A vanity metric is easy to increase and easy to celebrate, but unclear in meaning and weakly tied to value.

A quick test: is this metric actionable?

• Decision link: If the metric changes, do you know what decision to make?
• Controllability: Can your team influence it within a reasonable time?
• Value link: Does it correlate with revenue, retention, or pipeline in your context?
• Comparability: Can you compare it across channels/campaigns without misleading differences?

Examples: vanity vs actionable (and what to use instead)

Metric sets for common questions (keep it small)

Use “one primary + three supporting” as a default. Below are practical metric sets that translate common questions into measurable answers.

Question: “Should we increase spend on Search?”

• Primary: Cost per incremental conversion (or CPA if incrementality is not available)
• Supporting: Marginal ROAS, conversion rate, impression share (to see if you are constrained by budget vs. demand)

Question: “Which channel brings customers who stay?”

• Primary: 60/90-day retention by acquisition channel (or cohort LTV)
• Supporting: CAC by channel, payback period, churn/refund rate by channel

Question: “Are we spending too much on retargeting?”

• Primary: Incremental conversions from retargeting (lift vs. holdout, if possible)
• Supporting: Frequency, CPA vs. prospecting, overlap rate with organic/direct conversions

Question: “Which email campaign is worth sending again?”

• Primary: Revenue per recipient (or conversions per recipient)
• Supporting: Deliverability/open rate (diagnostic), click-to-conversion rate, unsubscribe/spam complaint rate

Question: “Is our top-of-funnel content helping?”

• Primary: Assisted conversions or influenced pipeline (defined consistently)
• Supporting: New users to qualified action rate (e.g., content → signup), return rate, time-to-next-step

Practical worksheet: turn a question into metrics in 5 minutes

1) Decision: ____________________________________________
2) Options (A/B/C): ______________________________________
3) Success outcome (business): ____________________________
4) Unit of analysis: ______________________________________
5) Primary metric (1): ____________________________________
6) Supporting metrics (2-3): ______________________________
7) Time window: __________________________________________
8) What action will we take if metric is up/down? __________

When you consistently start with the decision, map goals to outcomes, and choose actionable metrics over vanity metrics, marketing analytics becomes a daily operating tool rather than a reporting exercise.