Open handbook navigation

Metrics and analytics

North Star + input metrics

How to choose North Star Metric and Input Metrics: Good Score Criteria, Decomposition and Protection from Local Optimization

North Star should reflect the customer’s regularly acquired value and be linked to sustainable business outcomes. The team influences it through input metrics, but each link remains a hypothesis: check whether the main metric is growing at the expense of quality, margin or inappropriate segment.

What is North Star Metric?

Definition and substance

North Star Metric (NSM) is a metric that reflects the value of a product to users and the growth of the business itself. This is the key number that the entire team is targeting. For example, Spotify is the number of minutes you listen to music, for Airbnb it is the number of nights you stay.

Example of application

For an online educational platform, the logical choice is the number of successfully completed courses per period. This number shows both engagement and business benefits.

How to Choose North Star Metric

Principles of choice

Good NSM:

  • Reflects user value (what people stay and pay for)
  • closely related to revenue growth or core KPI
  • understandable to every employee

Anti-pattern: NSM should not be about vanity metrics like total registrations or downloads. These numbers do not guarantee growth or loyalty.

Cases of right and wrong choice

The food delivery service wants to grow and retain customers. The worst option for NSM is the number of new users. The good one is the number of successful orders completed per month by repeat customers. This indicator reflects both the quality and the real expansion of the business.

Input Metrics and their role

Definition and relationship with North Star

Input metrics are managed metrics that directly impact North Star Metric. While NSM looks like the end result, input metrics are the levers you can pull at every day.

Example:
For NSM, the “daily number of active taxi orders” input metrics will be:

  • average waiting time
  • share
  • percentage of drivers on the line

How to Find Input Metrics

Consider the user’s path. At every step, ask: What can be improved, measured, accelerated? All metrics that are systematically raised by NSM are candidates for input metrics.

Example: In music (NSM: minutes of listening), input will be the number of tracks started, frequency of playlist use, average time in a session.

Experiments: How to Use Metrics for Growth

Cycle of Experiments and Metrics

When you’re doing an A/B test or a product experiment, it’s important to choose what you’re measuring in advance. NSM’s growing – great. No, look what input metrics are moving.

Example:
In the hotel reservation service, the landing page is changed to search for a room. NSM is the number of nights of stay. Input metrics – Conversion to viewing details, adding to the cart. You see that input metrics grow after a change, but NSM doesn’t. So there’s a bottleneck in the other step.

Conclusions from experiments

The team’s task is to look not only at the NSM, but also to understand how input metrics have contributed. Sometimes you need a cascade: changed input metric, tracked the signal to NSM, concluded – repeated the cycle.

Mistakes and anti-patterns

Common mistakes

  • Non-working NSM (e.g. total number of downloads instead of DAU/WAU)
  • Ignoring input metrics – the team doesn’t know what to move
  • Re-optimize input metrics without final result control (NSM)
  • Output and action (input)

Example

SaaS product has increased the number of input metrics, hoping that NSM (paid subscriptions) will grow. In reality, the outflow increased due to annoying emails. Conclusion: Not every input metric is useful, you need to link to the final metric.

Conclusions: How to make metrics a working tool

  • North Star shows what the whole team is working on
  • Input metrics – levers for daily operating work
  • Each experiment should be measured in both categories of indicators.
  • People need to understand why NSM is chosen and what changes as input metrics grows.

For a deeper treatment, read the Amplitude guide to the North Star Metric and the Reforge series.


FAQ

What is the difference between North Star Metric and KPI?

NSM reflects the main value of the product to the user and business, and KPIs can be associated with any local goals (for example, the speed of updating the application).

Can I change the North Star Metric?

Yes, when your business model, audience, or strategy changes dramatically. NSM is usually stable and is reviewed no more than once a year.

How much input metrics should you have?

Usually 2-5 keys to manage them. Too much - focus is lost, little - no reason for the NSM to rise or fall is seen.

In what cases is input metric harmful?

If the growth of this input metric does not lead to the growth of the North Star. An example is the increase in the number of push notifications with a constant outflow of users.

What mistakes are most often made?

Select a surface NSM, do not control input metrics’ relationship to the result, do not review the set of metrics when a product or market changes.

Where to look for benchmarks for your niche?

General research on product analytics, reports of analytical platforms (for example, Amplitude, Mixpanel), in special guides on SaaS, e-commerce, media.