Identify product analytics core metrics
1. Write down the first part of your product's North Star vision (core metric).
This is what your entire company is aiming for-a statement of the vision of your product. Some examples, formulated for popular product companies: Help companies talk to their customers easier (Intercom) Allow low-income individuals to earn money by driving customers to a destination (Uber) Allow anyone to listen to undiscovered artists anywhere, anytime (SoundCloud)
2. Write down the second part of your product's North Star Metric, a single metric that tells you whether your current strategy is on-track like Daily% of refunded trips.
Examples: Monthly avg. conversation resolution time (Intercom) Daily % of refunded trips (Uber) Weekly % of registered users with at least 3 tracks played” (SoundCloud)
3. Define desirable outcomes that align to the "width" dimension of your North Star
Each dimension accounts for a different aspect of how this North Star is derived: Width Depth Frequency Efficiency Width allows you to measure the spread of your product. For example: Have all our customers logging in to our product Have none of our customers logging support requests
4. Define desirable outcomes that align to the "depth" dimension of your North Star
Depth makes sure you account for the details of your product. For example: Each customer purchases multiple items per order Each user listens to more than one track per session
5. Define desirable outcomes that align to the frequency dimension of your North Star
Frequency is how you drive higher usage of your product. For example: Each customer makes multiple rides using your app Each visitor clicks multiple ads when they visit your product
6. Define desirable outcomes that align to the "efficiency" dimension of your North Star
“Efficiency” is the way for you to spend your resources in the best way possible. For example: Each support case is resolved successfully All items from orders are delivered on time
7. From all of your outcomes from the previous steps, define 1-2 metrics for each dimension
A limit of 2 metrics for each dimension means you can still decide on which one you want to finally use, for example: Width: have all our customers logging in to our product Metric 1: % of customers that logged in within the last 7/30/60/90 days Metric 2: avg. Duration between logins per customer
8. Pick the final four metrics you will use and define the thresholds/variance
Each KPI will have a: threshold: value at which you trigger a business alert (e.g. 0 logins in the last 7 days) variance: how much variance you can expect this KPI to have over time (e.g. +5% MoM growth in logins) A stakeholder from each aspect of the business should be involved in this process to ensure total ownership.