TL;DR: These three frameworks are not alternatives to each other. They operate at different levels and answer different questions. A North Star Metric tells you which direction you are going. KPIs tell you whether the engine is running. OKRs tell you what must change before next quarter ends. You need all three, in that order, and confusing them is the most reliable way to produce metrics that feel meaningful but drive no decisions. This guide covers the precise definitions, the hierarchy between them, and the specific failure modes of each.

The definitions, stated clearly

North Star Metric (NSM): A single, enduring measurement that captures the core value your product delivers to customers. It is stable across quarters, expresses customer value rather than business extraction, and acts as a leading indicator of long-term revenue. It answers: are we going in the right direction?

Key Performance Indicators (KPIs): A small set of ongoing measurements that signal whether the business is healthy right now. They are continuous, do not expire, and live permanently on dashboards. They answer: is the engine running?

OKRs (Objectives and Key Results): A time-boxed goal-setting framework that defines what must measurably change in the next cycle, typically a quarter. They answer: what must improve, by how much, and by when?

The relationship between them: the North Star Metric anchors direction. KPIs monitor ongoing health. OKRs define what you are doing to improve health and move toward the North Star in the current period. They are not competing frameworks. They are different layers of the same system.

The hierarchy that most teams get wrong

Most product teams either use all three interchangeably or treat them as three optional choices. Neither is correct.

The correct hierarchy:

North Star Metric sits at the top. It changes rarely, perhaps once a year at most, and reflects a fundamental choice about what customer value the company is building toward. OKRs should ladder up to it. KPIs should be monitored as health signals that indicate whether the NSM is likely to move.

KPIs sit in the middle, as guardrails and signals. They run continuously. When a KPI drops significantly, it triggers an investigation or an OKR to fix it. When an OKR moves a KPI in the wrong direction, it signals a trade-off that needs PM judgment. KPIs are not goals in themselves. They are monitors.

OKRs sit at the bottom, as the change mechanism. They are time-boxed. They define specific, measurable movement in key results that the team commits to delivering in a quarter. A key result should be a KPI or a direct input to the NSM, not a task or activity.

The failure mode of most product teams: OKRs become task lists, KPIs become reporting theatre, and the North Star Metric either does not exist or is a vanity metric nobody can act on. The system degrades into "we shipped these things and here are some numbers" rather than "we are measurably moving in the right direction for the right reasons."

Side by side: the comparison that matters

What each one is NOT

These distinctions are where confusion most often lives.

A North Star Metric is NOT:

• A revenue metric (MRR, ARR, revenue growth). Revenue is a business outcome, not a customer value measure. Companies that mistake revenue for their North Star optimise for extraction at the expense of value delivery, which works until it doesn't.

• A vanity metric (total registered users, total downloads). These measure acquisition, not value delivered to customers who stayed.

• An OKR. An NSM does not expire. An OKR does.

• Multiple metrics. A dashboard of North Stars is a list of KPIs.

A KPI is NOT:

• A target. A KPI is a measurement, not a goal. "Monthly active users" is a KPI. "Increase monthly active users by 20% this quarter" is an OKR key result.

• A Key Result. A KR is a KPI that has been time-boxed with a specific target and committed to in an OKR cycle.

• Automatically strategic. Measuring things that are easy to measure rather than things that matter is how KPI dashboards become reporting theatres that nobody uses to make decisions.

An OKR is NOT:

• A task list. "Launch the new dashboard, write three case studies, reduce ticket response time" is a to-do list, not an OKR. Key results must be outcome measurements, not activities.

• A KPI. KPIs run continuously. OKRs expire at the end of the cycle whether or not they were achieved.

• Guaranteed to ladder up to the NSM. This is the most common OKR failure in product teams: OKRs that are internally consistent and measurable but disconnected from the strategic direction the NSM represents.

A B2B SaaS example that is not Spotify or Airbnb

Most guides use Spotify's "Time Spent Listening" or Airbnb's "Nights Booked" as North Star examples. These are useful, but they are B2C products with enormous data sets and well-understood user behaviours. Here is how the three-layer system works for a B2B product team.

Company: A project management tool for engineering teams. Core customer: engineering managers at 20 to 200-person companies.

North Star Metric: Number of engineering managers who create at least one cross-project view per week.

Why this NSM: It captures the specific customer value the product exists to deliver (cross-project visibility without manual aggregation). It is a leading indicator of retention because managers who use this feature renew at a significantly higher rate than those who do not. It is expressed in customer behaviour, not business outcome. It is specific enough to guide investment decisions: features that do not help engineering managers achieve this outcome deprioritise relative to those that do.

KPIs monitored continuously:

1. Week-over-week active engineering managers (health signal for the primary user type)

2. 30-day retention rate for engineering managers (churn signal, closely tied to NSM)

3. NPS from engineering manager cohort (satisfaction signal and churn predictor)

4. Mean time from account creation to first cross-project view created (onboarding efficiency signal)

Why these KPIs: They are continuous monitors of whether the product is delivering value to its strategic target. A drop in any of them triggers investigation. None of them is a goal in itself.

OKRs for Q3:

Objective: Help engineering managers achieve cross-project visibility faster in their first week.

Key Result 1: Reduce mean time from account creation to first cross-project view from 11 days to 5 days.

Key Result 2: Increase percentage of engineering managers who create a cross-project view in week 1 from 23% to 40%.

Key Result 3: Increase 30-day retention for engineering managers who create a cross-project view in week 1 from 71% to 80%.

Why these OKRs: Each key result is a KPI with a specific target and a time horizon. Together they drive the North Star Metric by removing the friction that is currently preventing engineering managers from reaching the valuable behaviour in their first week. They expire at the end of Q3. Whether or not they are achieved informs the Q4 strategy.

How they work together in a planning cycle

A well-functioning three-layer system operates like this across a year:

Annual (or when strategy changes): Validate or update the North Star Metric. Is it still the right measure of the core customer value the product delivers? Is it moving? If it is flat or declining, investigate before setting OKRs that assume it is correct.

Quarterly (OKR cycle): Identify which KPIs are most critical to improve this quarter and which inputs to the NSM are underperforming. Set OKRs with key results that are directly tied to moving those KPIs or NSM inputs. Approve OKRs by checking: does achieving these key results move the NSM? If not, the OKRs are misaligned.

Weekly or monthly (KPI review): Review KPIs against their baselines. Flag anomalies. Assess whether current OKR initiatives are producing leading signals of key result movement. Adjust priorities if a KPI is dropping faster than the OKR response can address.

End of quarter (OKR review): Score key results. For each key result that was not achieved, diagnose why. Was the initiative wrong? Was the measurement wrong? Was the key result the wrong lever for moving the NSM? Use the answers to set better OKRs next quarter.

The team that runs this cycle consistently has a compounding advantage: each quarter's OKR design is informed by what was learned in the previous one, and the North Star Metric provides a stable anchor that prevents quarterly OKRs from pulling the product in contradictory directions over time.

The most common failures of each framework

North Star Metric failures:

Choosing a business metric instead of a customer value metric. Revenue, ARR, and subscriber count are business outcomes. They are lagging indicators of customer value, not measures of it. A North Star built on a business metric optimises for the company's extraction rather than the customer's outcome, which eventually degrades retention as customers find alternatives that deliver more value for the same cost.

Choosing a metric that cannot be influenced by product decisions. If the product team cannot make a feature decision that would plausibly move the NSM, it is not a product North Star. It may be a company North Star, but it will not guide product investment.

Having multiple North Stars. When a team cannot choose a single NSM, it usually means the strategic question has not been answered: which customer type is the primary target, and what value does the product exist to deliver to them? Two North Stars means two strategies, which means no strategy.

KPI failures:

Too many KPIs. When every metric that can be measured gets added to the dashboard, the KPIs stop being key performance indicators and become a data dump. A team that tracks 25 KPIs prioritises all of them equally, which means it effectively prioritises none of them.

KPIs with no owner and no response protocol. A KPI that drops with no defined response protocol is a metric, not an indicator. If a KPI drops 15% and nobody knows what action to take, it is not functioning as a key performance indicator.

Optimising KPIs directly rather than understanding what drives them. The product that engineers its NPS score by surveying only its happiest customers has improved the metric without improving the customer experience. The metric is gameable. The underlying reality is not.

OKR failures:

Key Results that are activities rather than outcomes. "Launch the redesigned onboarding flow" is a task. "Increase the percentage of new users who complete onboarding in under five minutes from 34% to 60%" is a key result. The difference is that a task can be completed regardless of whether it achieves anything. A key result cannot.

OKRs that do not ladder up to the North Star. A team whose quarterly OKRs are internally logical and measurable but point in a different direction from the company's North Star is executing efficiently in the wrong direction.

Setting OKRs at 100% attainability. OKRs were originally designed as stretch goals. The expectation in Google's implementation, which popularised the framework, was that hitting 70% of an OKR represented strong performance. A team that consistently achieves 100% of its OKRs is not setting ambitious enough targets. This matters for product teams specifically because conservative OKRs protect the team from accountability while preventing the organisation from understanding what the product can actually achieve.

Frequently asked questions

What is the difference between OKRs and KPIs?
KPIs monitor ongoing business health continuously. They do not expire and they do not require an active change effort. OKRs define time-boxed goals for what must measurably change in the current cycle, typically a quarter. A KPI might be "monthly active users." The OKR key result is "increase monthly active users from 12,000 to 18,000 by end of Q3." The KPI measures continuously. The OKR commits the team to moving it within a specific window.

What is the difference between a North Star Metric and a KPI?
A North Star Metric is a single KPI chosen specifically because it captures the core value the product delivers to customers and acts as a leading indicator of long-term business health. All KPIs measure performance. The North Star Metric is the one KPI that, if it moves in the right direction over time, predicts that everything else will follow. KPIs are monitors. The North Star Metric is direction.

Can a Key Result in an OKR be a KPI?
Yes, and it should be. A Key Result is a KPI with a specific target and a time horizon attached. "Reduce churn rate from 4.2% to 2.8% by end of Q2" takes the KPI (churn rate) and turns it into an OKR key result by adding a target and a deadline. After Q2 ends, the churn rate reverts to being a KPI monitored continuously. The key result expires.

How do OKRs relate to the North Star Metric?
OKRs should be designed so that achieving their key results moves the North Star Metric. The test: if the team achieves 100% of all its OKRs this quarter, will the North Star Metric improve? If the answer is unclear or no, the OKRs are misaligned with the strategic direction. Achieving OKRs that do not move the NSM is efficient execution of the wrong work.

How many KPIs should a product team track?
Between 3 and 7. The word "key" exists for a reason. A product team that tracks more than 10 KPIs has stopped distinguishing between what matters and what is interesting to measure. Each KPI should have an owner, a defined baseline, and a defined response protocol: if this drops by X%, we do Y. Without that, a KPI is a metric, not an indicator.

Should product teams use all three frameworks?
Yes, because they answer different questions that all need to be answered. The North Star Metric is the strategic compass. KPIs are the operational monitors. OKRs are the quarterly change engine. A team with only OKRs changes without direction. A team with only KPIs monitors without acting. A team with only a North Star Metric has direction without a mechanism for getting there. The frameworks are complementary, not alternatives.

What is a good North Star Metric for a B2B SaaS product?
A good B2B SaaS North Star Metric captures the specific customer outcome the product exists to deliver, expressed in customer behaviour rather than business outcome. It should be a leading indicator of retention and expansion, measurable in product analytics, and actionable by product decisions. Examples: for a project management tool, "number of projects where a cross-functional team has viewed the same status update in a week." For a revenue forecasting tool, "number of forecasts submitted to leadership without a manual spreadsheet correction." For an engineering documentation tool, "number of PRs that had a linked spec reviewed before the PR was opened." The test: if this metric improves, do customers find the product more valuable? If yes, it qualifies as a North Star candidate.

Sources: Andy Grove, High Output Management (OKR origins, 1983); Google OKR guide (re.work); Sean Ellis, North Star Metric definition; Amplitude Leading vs Lagging Indicators (August 2026); OKR Mentors OKRs vs the World framework (July 2026); Koji North Star Metric Framework (June 2026). Last updated: August 2026 · meetsquad.ai