Mapping a Growth Model Around a North Star Metric

How to run it

1. 1

   Define the North Star metric

   Find the number that increments when you deliver value to a customer, not the number that increments when you get paid. Ask how people would naturally behave if they were getting a lot of value, and pick the behaviour that best reflects the mission.

   Pro tip If the exact measurement is thorny, agree the principle first and use rough proxies; a directionally accurate number is enough to steer the business.

   Watch out Revenue is easy to measure but is the wrong North Star, because it is downstream of value delivery and invites ideas that are orthogonal to customer value.

2. 2

   List the acquisition levers above it

   Name the handful of genuine sources of new audience or new customers and rank them. Most businesses find far fewer real sources than they expect once collaborations, cross-posting and other minor channels are honestly assessed.

   Pro tip If one lever is obviously dominant, say so out loud; a simple top of funnel is a gift, not a failure of analysis.

   Watch out Do not list channels you wish worked. Check whether the crossover traffic actually exists in the data.

3. 3

   Define the habituation rate

   Put a conversion rate between first exposure and loyal, repeat consumption, and work out what has to happen for someone to cross it. In software this is the aha moment; identify the equivalent event and the micro-metrics that drive it.

   Pro tip Interview people who became loyal in the last one to three months and ask what the first thing they saw was and why they did not bounce.

   Watch out Measure this on new users specifically. Metrics dominated by your existing loyal audience will not tell you whether you are welcoming newcomers well.

4. 4

   List the engagement and retention levers

   Name the content types or product experiences that keep existing customers coming back, and pick one high-level metric that tells you whether you are doing enough of them.

   Pro tip Look for implicit feedback loops, where volume raises visibility and data, which in turn raises volume, and draw them on the chart.

   Watch out Levers that grow the audience and levers that retain it are often different, and can even feel at odds. Model them separately.

5. 5

   Add a monetisation rate

   Express monetisation as revenue per North Star unit, covering every income stream that depends on the audience. Break it down later into the proportion monetised actively versus passively and the average value of each.

   Pro tip Segment out revenue only once acquisition channels exist that are independent of the core audience.

   Watch out Monetisation sitting downstream of the North Star is normal. Do not force every revenue stream into the North Star itself.

6. 6

   Add a small set of check metrics

   Choose a few numbers, such as refund rate or customer satisfaction, that you watch to make sure you are not breaking something while optimising. These are not drivers and are not to be optimised.

   Pro tip Keep the list deliberately short so the check metrics do not become a second scoreboard.

   Watch out Adding check metrics you never actually look at is worse than having none, because it creates false comfort.

7. 7

   Ask which ratio you could realistically 10x

   Go through the model and ask which conversion rate, audience number or monetisation rate could plausibly be multiplied ten times, and whether doing so would multiply the business. That comparison, not enthusiasm, decides where resources go.

   Pro tip Sanity-check the ambition against outliers in your category before betting on it.

   Watch out Some numbers cannot be 10x'd at all. Pursuing them anyway burns years.

In the wild

Common mistakes

From the transcript