Criteria for Good metrics
Actionable (2, page 143)
- Demonstrate clear cause and effect
- Understand how value was achieved (eg was it engineering or marketing)
- Blame culture when metrics go down is avoided
Accessible (2, page 143)
- Everyone can get them
- Allows metrics to guide as they are single source of truth
- “Metrics are people too”
- E.g. website hit is not as accessible as customer visiting site
Auditable (2, page 143)
- Data is credible to employees
- Can pot check the data with real people to verify
Iterating Metrics

Product Metrics Structures
Pirate or AARRR Metrics
Originally by David McClure

Metrics
- Acquisition (prospects visit from various channels/ users find your product)
- Activation (prospects convert to customers/ users have their first great experience)
- Retention (customers remain active/ user return to your product)
- Referral (Customers refer prospects/ users recommend)
- Revenue (customers make your business money/ users pay for your product)
Benefits
- Can calculate conversion through each step of the funnel (3, page 106)
Shortfalls
- Does not consider user satisfaction (3, page 106)
HEART Framework
- This framework is for a specific product or feature
- Happiness (how satisfied the user is with the product)
- Engagement (how the user interacts with the product)
- Adoption (same as activation in Pirate Metrics)
- Retention (same as Pirate Metrics)
- Task Success (how easy is it for the user to complete the task)
Specific Metric Details
Retention Parameters
Retention Curves (1, page 243)
- Days since first use does not start at 0 usually as this would be 100% and would alter the scale of the graph
- Can use cohort analysis (i.e. plotting the retention rates of different user cohorts (groups) onto the same axis to see the difference in the retention parameters for the separate groups

- Parameter 1 to notice: The percentage where the graph starts on Day 1 shows the initial drop off rate
- Parameter 2: Rate that the retention curve decreases from Day 1 value
- Parameter 3: Terminal value for retention curves is where the retention flattens out. If it is 0% then your product will ultimately lose all of its customers
Customer Life-time Value
Definition (4)
The amount of money made from a customer before that person switches to a competitor, stops using the product, or dies.
Effects of habits on CLTV (4)
- Some companies (like credit card companies) that have a high CLTV can justify spending an enormous amount of money on acquisition of new customers
- If people are in a habit of using a brand, that brand can safely increase their price and people won’t notice
- Games form habits and then offer people the option to pay a little to get another play as they are already hooked
Referral
Effects of habit on referral (4)
- Product growth by recommendation will be faster if the habit loop is shorter. I.e. if people come back every day there will be more recommendations, adding friends, and sharing
Acquisition
Overcoming Competitor Habit (4)
- Creating a product that is only marginally better than one with which users have already formed habit (e.g. the QWERTY keyboard is not the best but marginal improvements haven’t been successful)
- Dramatic improvements only can shake users out of habits
- People regress to engrained habits over time and new habits are lost (last In First Out LIFO)
- Requires more cognitive effort to switch to bing if you have always been using google even if they are identical
References
- Lean Product Playbook by Dan Olsen
- Lean Startup by Eric Ries
- Escaping the Build Trap by Melissa Perri
- Hooked by Nir Eyal
6 thoughts on “Product Metrics”