What should I measure first in my marketing?

Measure the business outcome you actually want, then work backward to the few signals that predict it. For most companies that means qualified leads or revenue first, with a small set of supporting metrics such as conversion rate and cost per qualified lead. Skip vanity numbers until the main outcome is tracked reliably. The what-to-measure-first page gives an ordered method for deciding which numbers earn a place on your dashboard.

How many metrics should I track?

Fewer than you think. A useful rule is that every metric on your dashboard should be tied to a decision you could make differently depending on the number. If a metric would not change your action whether it went up or down, it is probably noise. Most small teams do well with one outcome metric, a handful of leading indicators, and one or two diagnostic numbers they look at only when something changes.

Why do my analytics numbers not match my sales numbers?

Analytics tools and sales records measure different things in different ways, so a perfect match is unrealistic. Differences come from tracking gaps, consent choices, attribution rules, offline conversions and how each system defines a lead. The goal is not identical numbers but a stable, understood relationship between them. Document the known gaps, keep definitions consistent, and trust direction and trend more than any single figure. Treat unexplained, growing gaps as a signal to investigate the tracking.

Can I prove that marketing caused a sale?

Rarely with certainty. Most buyers touch several channels, and no attribution model captures all of them. What you can do is gather converging evidence: tracked sources, asking customers how they found you, and changes in results when you change effort. Be clear in reports about what is measured and what is inferred. Honest, imperfect attribution supports better decisions than a confident number that pretends to be exact.

Measurement

Measurement: tracking what changes a decision

Track the few numbers that change a decision, ignore the rest.

By Adnan GourijaHead of Search Marketing at Webtec · Reviewed 2026-10-17

Add us as a preferred source on Google

1 page

Measurement decisions determine what you track, what counts as success, and how you tell cause from coincidence. Good measurement is not about collecting more data. It is about choosing the few numbers that would change what you do next, and being honest about what they cannot tell you.

Founders usually err in one of two directions. They track everything and act on nothing, drowning in dashboards, or they track only the easy numbers, such as traffic or impressions, which feel like progress but may not connect to revenue. A third mistake is reading short-term movement as proof. Attribution is imperfect in every channel, and a page that admits that earns more trust than one that pretends otherwise.

This pillar has one page so far: what to measure first. It sets the order, from the business outcome you care about down to the leading indicators and tracking you need to see it. It pairs with the conversion pillar, because measurement tells you where to look, and with the systems pillar, because every channel decision needs a definition of working before you spend.

DecisionMeasurementWhat should you measure first, second and third in marketing?Measure in five layers, in order. First, one business outcome and a proof event that fires correctly. Second, which channels produce that outcome.…5 moves · 12 min

Questions about measurement

01
What should I measure first in my marketing?
Measure the business outcome you actually want, then work backward to the few signals that predict it. For most companies that means qualified leads or revenue first, with a small set of supporting metrics such as conversion rate and cost per qualified lead. Skip vanity numbers until the main outcome is tracked reliably. The what-to-measure-first page gives an ordered method for deciding which numbers earn a place on your dashboard.
02
How many metrics should I track?
Fewer than you think. A useful rule is that every metric on your dashboard should be tied to a decision you could make differently depending on the number. If a metric would not change your action whether it went up or down, it is probably noise. Most small teams do well with one outcome metric, a handful of leading indicators, and one or two diagnostic numbers they look at only when something changes.
03
Why do my analytics numbers not match my sales numbers?
Analytics tools and sales records measure different things in different ways, so a perfect match is unrealistic. Differences come from tracking gaps, consent choices, attribution rules, offline conversions and how each system defines a lead. The goal is not identical numbers but a stable, understood relationship between them. Document the known gaps, keep definitions consistent, and trust direction and trend more than any single figure. Treat unexplained, growing gaps as a signal to investigate the tracking.
04
Can I prove that marketing caused a sale?
Rarely with certainty. Most buyers touch several channels, and no attribution model captures all of them. What you can do is gather converging evidence: tracked sources, asking customers how they found you, and changes in results when you change effort. Be clear in reports about what is measured and what is inferred. Honest, imperfect attribution supports better decisions than a confident number that pretends to be exact.