Separate activity from value
Account creation, page views and notifications sent can grow while the customer experience gets worse. Ask what successful customers actually accomplish. A scheduling product might consider completed useful meetings; a workflow product might consider successful workflows that solve a recurring task.
The word “useful” needs an operational definition. Specify eligible customers, valid events, time period and exclusions. If one account can generate thousands of trivial events, the metric may reward intensity rather than broad customer value.
Evaluate candidate metrics
| Candidate test | What to examine |
|---|---|
| Value | Does the event represent an outcome the customer cares about? |
| Recurrence | Does repeated value matter for continued use? |
| Breadth | Can the team distinguish more customers receiving value from a few heavy users? |
| Reliability | Can the event be measured consistently? |
| Economics | Does improvement coexist with healthy retention and delivery cost? |
Connect the metric to inputs
- Define the value event in language that sales, product and analytics share.
- Identify acquisition quality, activation and recurring use as candidate inputs.
- Compare behavior across cohorts rather than assuming every event causes retention.
- Add guardrails for customer satisfaction, cost and harmful behavior.
- Revisit the definition when the product or business model changes.
A correlation with retained customers is a reason to investigate, not proof that forcing the event will improve retention. Customers may perform the event because the product already fits their needs.
Worked example: distinguish breadth and frequency
Illustrative example: an automation product reports 50,000 successful runs. That number could represent 500 businesses receiving regular value or five accounts running intensive background tasks. Report active value-receiving accounts alongside runs per account. The team can then see whether growth reaches more customers or depends on a small group.
If runs increase because a workflow repeats unnecessarily, the apparent North Star improvement might increase infrastructure cost without delivering value. An operational quality guardrail prevents that misleading conclusion.
Avoid making the metric a target in isolation
People adapt to targets. If the team is rewarded only for event volume, it may add unnecessary notifications or low-value steps. Keep the customer outcome visible in reviews and inspect unusual changes in the metric's composition.
Can revenue be the North Star?
It can guide the business, but revenue alone often says little about which customer behavior creates recurring value. Pair business outcomes with operational inputs the team can influence.
Do all products need one metric?
Not always. A marketplace or a business with several distinct products may need a small set of coordinated measures rather than one oversimplified number.
Put this into practice
Write your proposed metric as a complete definition, then describe two ways the number could rise without improving customer value. Those scenarios tell you which guardrails you need.
Related foundation: SaaS growth strategy: build acquisition around activation. How these guides are prepared.
Related portfolio work: eGrow. The worked examples in this guide are illustrative and are separate from the portfolio’s project evidence.
