North Star Framework: The Metric, the Inputs, and the Work

Created by Amplitude (John Cutler and Jason Scherschligt) - https://amplitude.com/north-star-hub

Overview

The North Star Framework is a product management framework built around one number. Amplitude's North Star Playbook defines it as "a product management model based on a single metric," the North Star Metric, that best captures the value customers derive from a product. The framework adds two layers under that metric: a small set of inputs that together produce it, and "the work," the research, design, engineering and testing that teams do every day to move the inputs. A workshop is how most teams build the first version. This page covers the framework as an operating model. The metric on its own, and how to choose and decompose it, has its own page on the North Star Metric.

The term North Star Metric has its own origin. Stream's explainer credits Sean Ellis, founder of GrowthHackers, with coining it and quotes his definition: "the single metric that best captures the core value that your product delivers to customers." Amplitude's playbook builds a full operating model around the metric. Amplitude published The North Star Playbook in December 2019, and co-author Jason Scherschligt wrote at the time that he had written it with John Cutler and that Amplitude sponsored the project. Amplitude's North Star hub calls Cutler "the original author of the North Star Playbook." The playbook itself refers to "Amplitude's North Star Framework," so the creator line credits Amplitude and both authors.

The playbook sets three qualities for the metric. It represents the value users get from the product, it sits within product and marketing's sphere of influence, and it is a leading indicator of revenue. It then lists three purposes for the framework as a whole: it helps prioritize and speed up informed but decentralized decision-making, it helps teams align and communicate, and it lets teams focus on impact and sustainable, product-led growth. Product and marketing teams may pull on very different inputs, but they check their choices against the same North Star.

The structure is a tree. The playbook describes its diagram as "a scaffold containing assumptions, beliefs, and causal relationships." At the bottom sits the work, above it the inputs, above them the North Star Metric, and at the top the medium- and long-term business results it is expected to lead. The playbook calls inputs independent variables and the North Star an outcome, a dependent variable. In Cutler's words from the playbook, "The goal of the North Star is to be one level out of reach." Teams move inputs, and the North Star shows whether those moves added up.

Cutler is clear that the tree is not a chain of command. In a 2021 post on cascades he writes that models like the North Star Framework "are not cascades" because there is no hierarchical relationship between the metric and the inputs. The relationship is causal and leading or lagging. Goals such as OKRs can still be set, and in his framing they map the work to the inputs. That difference matters in practice: a cascade hands each level a target from above, while the framework asks each team which input its work can move and why.

The playbook also lists what the framework is not. It is not a roadmap, a software development process, a prioritization framework, a goal-setting system such as OKRs, management by objectives, or a one-time fix. It informs all of those. Its chapters on roadmaps, prioritization and OKRs show how to connect them, for example by framing OKRs as the expected impact of shorter bets on one or more inputs.

Most of the value comes from the conversations the model forces. The playbook's closing chapter says teams leave with "nagging disagreements" and still have meaningful conversations about beliefs, value exchanges, inputs and outputs, and bets. It sums this up in one line: "The quality of your conversations is the real secret to the North Star." A team that fills in the boxes without having those conversations ends up with a metric on a slide and nothing else changed.

Core Principles

Connect the Three Languages

The playbook opens with a product manager who found her company speaking three languages: the customer's (needs, goals, experiences), the product's (features, workflows, releases) and the business's (vision, differentiation, revenue, growth). The framework exists to connect them. The North Star Metric expresses customer value in a way the business can read as a leading indicator, and the inputs express it in a way product teams can act on. If a draft North Star makes sense to only one of the three groups, keep working on it.

Keep the Metric One Level Out of Reach

The North Star should not be something a team can move directly. The playbook says never to try to influence it directly and to work on the inputs instead. Cutler describes the metric as a composite of the parts of the company working together, meant to get people asking why it went up or down. A metric one team can push on its own measures that team's output, and the rest of the organization stops caring about it.

Treat the Tree as a Set of Beliefs

Every link in the tree is a bet that one thing causes another. The playbook calls the first pass at a North Star "literally a hypothesis about the levers of growth" and expects teams to go through a cycle of doubt before they converge. It asks teams to write down beliefs about customers, competitors, technology and causation, because unspoken beliefs leave unresolved tensions in the team. Writing them down turns a disagreement about the metric into a disagreement about a specific assumption, which data can settle.

Put Words Before Numbers

Teams tend to jump straight to formulas. Cutler's statement exercise starts with a sentence about the path to sustainable growth and treats measurement options as the last step. The playbook's reason is blunt: if someone cannot explain the North Star qualitatively, they will not be able to explain it quantitatively. It also warns against dropping promising ideas just because nobody yet knows how to measure them.

Decide Locally Against a Shared Reference

The framework is meant to let teams make decisions without escalating every trade-off. Each team chooses which input to work on and how, and every team checks those choices against the same North Star. That only works if there is one North Star per product and customer base; the playbook treats the urge for several as a common trap and reserves multiple North Stars for distinct lines of business with different customers. With one reference, a dispute between two teams becomes a comparison of expected impact.

Make It Part of the Operating Rhythm

A workshop output that nobody revisits changes nothing. The playbook lists what teams that stuck with the framework had in place: a sponsor with influence and authority, leadership buy-in, communication and change management, onboarding for new team members, and approval processes. It also lists signals that the framework is working, such as team members explaining how their daily work connects to the North Star and people outside product using words like "inputs."

Steps

1. Name the game and the value exchanges Start by deciding which game the product plays. The playbook describes three: an attention game, where more time in the product signals satisfaction; a transaction game, where users want the right product and an easy transaction; and a productivity game, where users have a job to do. Then list the key value exchanges, the moments where a customer's time, attention or money is rewarded. The playbook says most products can narrow these to a handful and that some happen outside the product. The output is one sentence naming the game and a short list of value exchanges.

2. Write the North Star statement Before choosing any metric, complete the playbook's statement worksheet: our path to sustainable growth is a function of our ability to do something, which is a function of a few inputs, which the team can influence through named opportunities and interventions. Cutler says teams have spent entire workshops "just chipping away at this statement." Surface beliefs and review the product vision while you write it. Only when the words hold up do you list measurement options under each box.

3. Run the workshop Bring a small cross-functional group into a focused session. Amplitude's workshop guide runs it in about two hours: an opening discussion, the game, the checklist of a good North Star, the statement exercise, the structure of metric and inputs, a warm-up on another company's product, then silent drafting and convergence. The output is a few strong candidates for the metric and its inputs, plus the questions still open. The workshop skill covers facilitation in detail.

4. Define the metric and inputs precisely Give the metric and every input a name and a definition, using the playbook's template: "Our North Star Metric is called X, which we define as Y." Test the metric against the seven-item checklist and look at how it would have moved in historical data. Test the inputs by asking whether a team can generate ideas to influence each one and whether current roadmap items map to them. Expect several rounds; the guide's closing questions ask what additional information you need before deciding.

5. Connect the work through levels of bets Link every piece of work to an input. The playbook's levels of bets model runs from Level 0, which ties the North Star to company success over years, to Level 3, which covers interventions a team can finish in a sprint. Put the bets on a roadmap that shows at a glance how in-progress and planned work connect to the North Star. Review each bet when it finishes instead of marking it done and moving on.

6. Make it stick Name a sponsor, get leadership to use the metric in their own reviews, and add the North Star to onboarding. Amplitude reports its own North Star weekly in product and leadership teams and quarterly at company All Hands, according to the playbook. Share the work repeatedly in formal and informal forums. Watch for the playbook's signals, such as people saying "no" more easily and with more evidence.

7. Review and revise the model Check the tree on a regular rhythm. Ask whether your confidence in the inputs and the North Star has gone up or down in the last couple of weeks, and why. Revise inputs and definitions as you learn. Revise the North Star itself when the strategy shifts or, as Amplitude's Ted Clark puts it in the playbook, when it stops directionally indicating where revenue is going.

The Four Layers of the Framework

The playbook's statement worksheet and its levels of bets model describe the same four layers from two angles: what each layer should look like, and how long a bet at that level takes to show results. Each row cites the North Star Playbook.

The playbook notes that time horizons vary by company, and that the higher-level bets, Levels 0 and 1, are the ones most closely tied to the North Star Metric and its inputs. Reading the table from the bottom up shows how a sprint-sized change is supposed to connect, through an opportunity and an input, to the metric.

When to Use

• Several teams ship into one product and each reports a different measure of success. The framework gives them one outcome to check decisions against and separate inputs to work on.
• Strategy discussions keep producing agreement that fades within weeks. The workshop and statement exercise put the strategy into words and a structure the team can refer back to.
• The roadmap is a list of features and nobody can say which outcome each one serves. Linking work to inputs through levels of bets exposes items with no theory of impact.
• The company is moving toward product-led growth. The playbook says the framework works especially well in companies that use a product-led growth model, which it describes as using the product, rather than marketing or sales, to drive acquisition, retention and monetization.
• Leadership sees revenue and pipeline but no early signal of whether customers are getting value. A leading metric with inputs underneath gives that signal before it shows up in revenue.

When Not to Use

• The business is really several businesses with separate customers and separate profit and loss statements. Build one framework per line of business instead of forcing a single tree over all of them.
• Nobody with authority will sponsor it. The playbook lists a sponsor and leadership buy-in among the things successful teams had, and without them the North Star becomes one more chart on a wall.
• The team wants a quick fix for a delivery problem. The framework is not a development process or a one-time fix, and it will not repair missed deadlines or unclear requirements.
• The main goal is a target to tie individual bonuses to. The framework is built on hypotheses that are expected to change, and hard incentives on one number invite people to move the number without moving value.