8 B2B SaaS North Star Metrics to Track in 2026
Stop Waiting for Data. Find Your North Star Now.
Your gut says growth is slowing, but the numbers are stuck behind a queue. You file a BI ticket, wait, get a dashboard later, and it still doesn't answer the follow-up question you care about. Waiting on data is a business killer. In 2026, you shouldn't need SQL, a sprint cycle, or a patient analyst just to understand what's happening in your own SaaS business.
That's why a North Star Metric matters. It gives the team one number that reflects real customer value and business progress. In B2B SaaS, the idea became mainstream when Amplitude pushed the concept into the market in 2016, and Finmark's compilation notes that more than 80 companies adopted specific North Star Metrics after that wave of thinking took hold, as summarized in CXL's review of North Star Metrics.
The hard part isn't understanding the concept. It's choosing the right metric for your model, stage, and motion. A PLG company shouldn't blindly copy a sales-led CRM. An AI analytics product shouldn't obsess over logins if the actual value is insight generation. Good North Stars focus teams. Bad ones create fancy dashboards and bad decisions.
This B2B SaaS North Star Metrics List gets to the point. You'll see which metrics fit which business model, where they work, where they mislead, and how to track them with Statspresso, a Conversational AI Data Analyst that lets you skip the SQL and just ask your data a question.
1. Monthly Recurring Revenue
A founder checks the board deck and sees MRR up. Good news, until the next question lands. Is growth coming from new customers, expansions, annual contracts normalized monthly, or a pricing change that won't repeat? If MRR is going to be your North Star, it has to explain the business, not just decorate the slide.
MRR earns its place because it ties product, sales, pricing, and retention into one operating number. For sales-led SaaS, that often makes it the right company-level metric once the team has a repeatable motion. For PLG SaaS, MRR is usually better as the top-line outcome, with activation, adoption, or free-to-paid conversion acting as the leading indicators underneath. Early-stage teams get into trouble when they force MRR to answer a product-market-fit question it cannot answer.
That trade-off matters. A mature sales-led company can run the business off MRR slices by segment, rep, and expansion path. A PLG company still needs to know what user behavior creates revenue later. If self-serve users never reach the core value moment, MRR will lag the problem by weeks or months.
When MRR works and when it doesn't
MRR works best when pricing is subscription-based, reporting is clean, and the team already knows the funnel stages that produce durable revenue.
Use it well:
Segment MRR: Break it out by plan, industry, sales-led vs. self-serve, or integration type.
Split new, expansion, contraction, and reactivation MRR: A single growth number hides too much.
Read it with product metrics: Pair MRR with activation, churn, and feature adoption so teams know what to fix.
Common mistakes:
Using one headline number for every decision: Finance can live with that. Operators cannot.
Treating all revenue as equal: Enterprise expansion and SMB new business behave differently.
Reviewing it only monthly: Weekly checkpoints catch changes in pipeline conversion and expansion earlier.
Practical rule: MRR is a strong scoreboard. It still needs leading indicators if you want to steer.
For teams using Statspresso, this metric gets more useful once you ask follow-up questions in plain English instead of waiting for a custom dashboard. Ask for MRR by acquisition motion, by workspace size, or by first integration connected. That is where a conversational AI data analyst helps. Founders and PMs can inspect revenue patterns day to day without handing every question to a data team.
A good setup starts with clean definitions. Decide whether discounts, services, annual prepaids, and paused accounts belong in reported MRR. If the team uses different logic across finance, product, and GTM, the metric becomes a source of arguments instead of decisions. For a useful primer, Statspresso's guide to understanding key performance indicators pairs well with this section, and this breakdown of calculating MRR for SaaS founders is a solid reference for the finance side.
Try asking Statspresso: “Show monthly recurring revenue for the last 12 months, split into new, expansion, contraction, and churn MRR, and compare self-serve vs. sales-led customers.”
2. Customer Acquisition Cost
A familiar SaaS problem: revenue is growing, the pipeline looks full, and the board still asks why cash burn is not improving. CAC usually explains the gap. If winning each new customer costs too much, growth adds pressure instead of margin.
Customer Acquisition Cost is the clearest test of whether your go-to-market motion scales. In an early search phase, CAC jumps around because the team is trying channels, pricing, and messaging. Once a motion starts repeating, CAC becomes a discipline metric. It shows whether demand gen, sales, onboarding, and product handoff work as one system or as four separate teams creating friction for each other.
CAC also changes meaning by business model.
For PLG SaaS, I care less about one blended number and more about CAC by channel and activation path. Paid social may generate signups. Organic search may bring in buyers with a specific reporting problem and a faster path to value. If the product gets users to a useful first output quickly, paid acquisition can work. If activation is weak, even cheap traffic becomes expensive.
For sales-led SaaS, segment-level CAC matters more than top-line averages. SMB, mid-market, and enterprise deals carry different sales cycles, implementation costs, and rep involvement. A blended CAC can hide a bad enterprise motion or make an efficient SMB engine look worse than it is. Hybrid companies need even more honesty. A “self-serve” funnel that depends on SDR rescue calls is not self-serve in cost terms.
That is why PMs should care about CAC, not just finance and growth leads. Friction in signup, weak onboarding, slow time to first value, and poor product-qualified lead signals all raise acquisition cost. CAC is partly a marketing number and partly a product operations number.
A few patterns are worth tracking closely:
PLG teams: Measure CAC by source, signup cohort, and activation milestone.
Sales-led teams: Measure CAC by segment, sales team, and payback period.
Hybrid teams: Separate true self-serve CAC from sales-assisted CAC.
Growth-stage teams: Watch trend lines weekly, not just monthly, because channel efficiency can deteriorate fast.
Cheap CAC from bad-fit customers still produces expensive growth later.
I rarely recommend CAC as the single North Star metric. It works better as a guardrail around growth because teams can improve CAC the wrong way by cutting spend too aggressively, narrowing ICP too far, or underinvesting in sales capacity. The better question is whether CAC is justified by retention, expansion, and payback for the customer type you are acquiring.
For day-to-day decision-making, a conversational AI analyst is often more useful than another static dashboard. In Statspresso, founders and PMs can ask for CAC by channel, by segment, by sales-assisted vs. self-serve, or by the first data source connected after signup. That last cut is especially useful in analytics products because it links acquisition cost to buyer intent, not just lead source. If renewal economics are part of the picture, teams can pair CAC analysis with the WhatPulse sales renewal tool to pressure-test whether acquisition costs still make sense after renewal and expansion assumptions.
Try asking Statspresso: “Show CAC for the last two quarters by acquisition channel, split self-serve vs. sales-assisted, and compare activated customers vs. non-activated signups.”
3. Net Revenue Retention
A company can post strong new bookings and still have a retention problem. Net Revenue Retention shows whether the installed base is getting more valuable over time.
That makes NRR one of the clearest strategic metrics in B2B SaaS, especially once the business has enough cohort history to trust the signal. It answers a tougher question than topline growth. Are existing customers renewing, expanding, and finding more reasons to stay?
For sales-led SaaS, NRR often reflects onboarding quality, account management, pricing design, and how well expansion paths match the buyer's rollout motion. For PLG companies, it usually rises when more teams activate, invite coworkers, connect more data, or adopt higher-usage workflows. Same metric, different engine.
What NRR reveals that MRR hides
MRR can increase while account quality gets worse. A busy sales team can mask poor retention for a while. NRR removes that cover because it isolates what happens after the initial sale.
Use NRR when:
You have real expansion mechanics: More seats, usage-based pricing, add-ons, extra modules, or multi-team rollout.
Customer value compounds over time: The product gets harder to replace after adoption spreads.
Product, success, and sales all shape outcomes: NRR captures the handoff between promise and delivered value.
NRR is less useful as the primary North Star when:
You are still very early: Small cohorts can swing wildly and create false confidence.
Pricing is mostly flat: If accounts have little room to grow, NRR behaves more like a churn metric.
Contracts are short and transactional: Retention noise can drown out the operating signal.
The practical move is to segment it. SMB, mid-market, enterprise, self-serve, and sales-assisted accounts each expand for different reasons. If one segment has high logo retention but weak expansion, that usually points to packaging or adoption gaps. If another expands well but churns after the first renewal, the problem is often implementation quality or a bad-fit ICP.
I also would not read NRR as a finance-only metric. It is one of the fastest ways to see whether your product's value moment turns into account growth. Teams that improve NRR usually do it by shortening time-to-value, increasing feature adoption in the first 60 to 90 days, and giving customers a clear reason to broaden usage inside the account.
That is why conversational analysis helps here. In Statspresso, a founder, PM, or CS leader can ask for NRR by cohort, segment, contract type, plan, owner, or activation milestone without waiting on a custom SQL pull. That matters most for companies without a dedicated data team, because the useful questions change every week. Which accounts expanded after adopting a specific feature? Which onboarding path correlates with stronger renewal? Which segment has healthy gross retention but weak net retention?
If your team is trying to improve renewals and expansion at the same time, the WhatPulse sales renewal tool is a practical way to model the commercial side alongside your NRR analysis.
Try asking Statspresso: “Show quarterly NRR for the last 8 quarters by segment and sales motion. Then identify the product actions shared by accounts that expanded in the last quarter.”
4. Customer Lifetime Value
A founder looks at CAC and feels good. The deals are closing, signups are up, and the pipeline looks healthy. Six months later, the same company realizes the new customers either never expanded, needed too much support, or left before payback. That is the problem LTV is supposed to catch.
Customer Lifetime Value measures how much economic value a customer produces across the full relationship. In B2B SaaS, that makes it a strategy metric, not just a finance metric. It helps answer a harder question than “Can we acquire customers?” It answers “Which customers are worth acquiring, keeping, and expanding?”
LTV gets more useful as the business matures. Early-stage teams usually do not have enough retention history to model it with confidence, especially in sales-led SaaS with annual contracts. In PLG, you can often read the signal sooner because activation, expansion, and early churn show up faster. Even then, I would treat early LTV as directional until the cohorts have aged.
A healthy LTV:CAC relationship matters, but the exact ratio depends on sales motion, gross margin, contract length, and cost to serve. A self-serve PLG product can tolerate a different profile than an enterprise product with long onboarding and heavy support. The practical rule is simple. If a segment looks expensive to acquire and slow to expand, its LTV needs to justify that burden.
The teams that use LTV well rarely look at one blended number. They break it apart so it can guide decisions:
By acquisition source: Organic, partner, outbound, and paid channels often produce very different customer durability.
By sales motion: PLG users who activate on their own behave differently from sales-assisted accounts.
By plan or contract type: Monthly self-serve customers and annual enterprise customers should not sit in the same LTV bucket.
By use case: Accounts that embed the product into a recurring workflow usually stay longer and expand more.
For a product like Statspresso, I would pay close attention to accounts that connect multiple data sources, return weekly, and share answers across functions. Those behaviors usually indicate the product moved from occasional reporting help to an operating habit. That is the kind of pattern that raises LTV in a durable way.
There are trade-offs here. Revenue-based LTV is fast to calculate, but it can mislead if onboarding, support, and success costs vary a lot by segment. Fully loaded LTV is more useful for decision-making, but it takes more discipline to maintain. The wrong move is using a single company-wide average and calling it strategy. That number hides both your best-fit customers and the segments draining time.
Statspresso is useful here because LTV questions change constantly, and smaller teams rarely have an analyst available for every follow-up. A founder can ask which 2024 cohorts have the highest projected LTV, whether sales-led accounts outperform PLG after 12 months, or which activation events correlate with longer retention. That lets teams investigate day-to-day without building a queue of SQL requests.
LTV is a check on acquisition quality, retention quality, and expansion quality at the same time.
Try asking Statspresso: “Compare projected LTV by acquisition source, plan, and sales motion for customers acquired in 2024 and 2025. Then show which activation behaviors are most common in the highest-LTV cohorts.”
5. Churn Rate Customer and Revenue
Monday morning looks great until the renewal report lands. New bookings are up, trial volume is healthy, and the product team is celebrating engagement. Then three good accounts cancel and two larger customers downgrade. Growth did not disappear. It was offset.
That is why churn belongs on any serious North Star shortlist. It tests whether growth is durable or just loud. I look at customer churn and revenue churn together because they answer different operating questions. Customer churn shows logo loss. Revenue churn shows whether the accounts leaving matter to the business.
As noted earlier, strong SaaS companies keep a tight grip on churn. The useful takeaway is not a universal benchmark. It is the operating habit behind it. Teams that review churn by segment, cohort, and product behavior catch weak retention earlier and make better decisions about acquisition, onboarding, and pricing.
Customer churn and revenue churn answer different questions
A PLG company can tolerate more logo churn if low-value self-serve users leave while strong-fit accounts activate, adopt, and expand. A sales-led company has the opposite risk. Logo counts can look stable while revenue churn creeps up because a handful of larger customers cut seats, downgrade plans, or fail to renew.
This is why a single churn number is not enough.
Track churn at least three ways:
By plan: Self-serve, mid-market, and enterprise tiers rarely churn for the same reason.
By company size: SMB churn often reflects weak onboarding or low urgency. Enterprise churn often points to rollout failure, missing governance, or poor executive buy-in.
By activation or usage depth: Teams that never hit the core value moment should not be mixed in with accounts that have reached full adoption.
Cohorts matter just as much. If customers acquired in one quarter churn faster than the rest, I would first inspect channel mix, onboarding changes, pricing updates, and product releases from that period. Without cohort analysis, churn stays abstract. With cohorts, it becomes diagnosable.
The trade-off is straightforward. Customer churn is easier to explain to the whole company. Revenue churn is better for prioritization because it tells you where money is leaking. Early-stage PLG teams often start with logo churn because volume is higher and contracts are smaller. Later-stage and sales-led teams should put more weight on revenue churn, gross and net, because one account can move the entire month.
Statspresso helps because churn questions rarely stop at one chart. A founder or PM can ask which cohorts have the worst six-month revenue churn, whether sales-assisted accounts retain better than pure PLG signups, or whether teams that connect a second data source are less likely to cancel. That kind of day-to-day analysis usually gets stuck in a BI queue. It should not.
Try asking Statspresso: “Show monthly customer churn and revenue churn for the past 12 months, broken out by plan and acquisition source. Then compare churn for accounts that reached activation in their first 30 days versus those that did not.”
6. Feature Adoption and Product Engagement
Many SaaS teams should start here, even if they don't. If you haven't nailed what real product value looks like, feature adoption and engagement are often better North Star candidates than revenue.
A North Star should reflect customer value, not just attention. That's why generic logins and pageviews are weak choices. The right engagement metric captures the moment the product becomes useful.
A strong B2B SaaS example comes from Dropbox. Finmark's compilation notes that Dropbox used “trial accounts with more than 3 active users in week 1,” and that metric predicted conversion to paid subscriptions at 40% to 60% based on internal benchmarks shared in Amplitude case studies, as summarized in CXL's North Star analysis. That's a great North Star because it captures collaborative value early, not passive traffic.
Good engagement metrics are specific
For analytics tools, “active users” is often too vague. A user can open the app, poke around, and leave without getting value. A better metric is tied to the useful job.
For Statspresso, better candidates include:
AI queries resolved per active workspace
Insights added to dashboards
Shared dashboards viewed by teams
Connected sources actively queried
The verified brief explicitly points to a gap in the market. Existing North Star content covers generic SaaS examples but rarely gives AI-specific guidance for products where value comes from generated insights instead of simple activity, as described in Growth Academy's North Star examples article. That gap is real. AI products need value metrics, not vanity traffic metrics.
What works here is choosing one action that means the user got something useful. What doesn't work is bundling every product event into an “engagement score” nobody trusts.
If your engagement metric goes up while retention stays weak, the metric is too shallow.
Try asking Statspresso: “What percentage of active users have adopted the AI Insight Gallery feature? Show me the trend over the last 6 months.”
7. Free-to-Paid Conversion Rate
A familiar PLG problem. Signups look healthy, the top of funnel keeps growing, and revenue still feels stuck.
Free-to-paid conversion cuts through that noise. It shows whether the product gets users to a buying moment, not just an account-creation moment.
This metric matters most for freemium, free trial, and hybrid SaaS models where users can experience value before talking to sales. In a sales-led company, it usually belongs lower in the dashboard. Procurement, security review, and contract timing can delay purchase long after the product has proved its value, so conversion inside the app is not always the cleanest company-level North Star.
What improves conversion is activation. Users pay after they reach a clear value milestone and understand why the product belongs in their workflow. As noted earlier, teams that anchor on first-value behavior usually make better decisions than teams staring at raw activity counts.
For Statspresso, I would track trial users who:
connect a real data source
ask several real business questions
save or share an output with a teammate
come back after the first session
Those actions show buying intent far better than logins or time in app. A user can spend twenty minutes clicking around and still be nowhere near a purchase decision.
There is a trade-off here. Push upgrade prompts too early and conversion can rise for a month while retention gets worse a quarter later. Wait too long and qualified users stay on free plans that never turn into revenue. The right threshold depends on the model. PLG products usually convert best right after a user gets a concrete result. Sales-led products often use product activity as a qualification signal for sales, not the final conversion event itself.
Use free-to-paid conversion as a North Star if the product does the selling. Treat it as a supporting metric if sales, legal review, or multi-stakeholder buying committees do the selling.
Try asking Statspresso: “What's our trial-to-paid conversion rate for cohorts who signed up in the last 3 months? Break it down by traffic source, and compare users who connected a data source in week one vs. those who didn't.”
8. Customer Health Score and Expansion Revenue Potential
A customer looks fine at renewal prep. Usage is still there. The champion still shows up to calls. Then procurement pushes the deal into a downsize, or a competitor gets invited into the account. That usually happens because the team tracked activity, but missed account health.
Customer health score is a working model for retention and expansion. It pulls together the signals that matter across product, success, support, and commercial teams. Used well, it answers two practical questions: which accounts need intervention now, and which accounts have earned an upsell conversation.
This metric is especially useful once a SaaS business has enough customers that account reviews become inconsistent. Earlier-stage PLG companies often use lightweight health logic to spot accounts that are ready for sales assist. Sales-led teams usually need a richer score because renewals, stakeholder adoption, and rollout depth matter more than raw logins.
The trade-off is clarity versus accuracy. Add too many inputs and the score turns into a black box that nobody trusts. Keep it too simple and it misses real risk. The best version is explainable in one minute by a CSM, PM, or founder.
For Statspresso, I'd build the score from account-level signals such as:
question volume over the last 30 days
number of active users per workspace
dashboards created, saved, or shared
connector setup depth across real business data sources
support tickets by type and resolution pattern
executive or cross-functional usage, not just one analyst
contract tier, renewal date, and recent seat changes
Then map the score to action. High-health accounts with broad usage and growing team adoption go to expansion review. Mid-health accounts get targeted enablement. Low-health accounts need root-cause analysis, usually around setup gaps, weak rollout, or low value realization.
Expansion potential should not be treated as the same thing as health. Some accounts are healthy but fully saturated on their current plan. Others have uneven health and still hold upside if one department is getting strong value while the rest of the company has not adopted yet. That distinction matters more in sales-led SaaS, where expansion often depends on stakeholder coverage and procurement timing, not just product engagement.
For founders without a data team, this is exactly the kind of metric that should be easy to query in plain English instead of buried in a CRM plus BI stack. Statspresso can pull product usage, billing context, and support signals into one view so teams can ask follow-up questions on the fly, without waiting on an analyst to rebuild a dashboard.
A useful health score changes who gets attention this week and why.
Try asking Statspresso: “Show me customers with high product engagement, low support friction, renewal in the next 120 days, and unused seat or workspace expansion potential. Group them by plan and flag the best upsell candidates.”
B2B SaaS: 8 North Star Metrics Comparison
Metric | Implementation complexity | Resource requirements | Expected outcomes | Ideal use cases | Key advantages |
|---|---|---|---|---|---|
Monthly Recurring Revenue (MRR) | Low, simple subscription math | Billing & subscription data, basic dashboards | Predictable monthly revenue and growth trends | Regular revenue monitoring, forecasting, investor reporting | Easy to communicate, central for valuation and forecasts |
Customer Acquisition Cost (CAC) | Medium, cost allocation & attribution required | Marketing & sales spend, CRM, attribution tooling | Cost per new customer and channel efficiency insights | Optimizing acquisition spend and go-to-market strategy | Reveals channel ROI; informs budget and payback analysis |
Net Revenue Retention (NRR) | High, track expansions, contractions & churn by cohort | Detailed MRR movements, billing changes, cohort analytics | Revenue retention including expansion; indicates sustainability (>100%) | Mature SaaS growth assessment, upsell effectiveness, investor evaluation | Shows expansion power; reduces reliance on new acquisition |
Customer Lifetime Value (LTV) | Medium, needs churn, ARPU and margin modeling | ARPU, churn rates, gross margin, historical cohorts | Estimated long-term revenue/profit per customer | Setting CAC targets, pricing strategy, segment prioritization | Guides acquisition spend; highlights high-value segments |
Churn Rate (Customer & Revenue) | Low–Medium, straightforward calc, needs segmentation | Customer records, revenue tracking, cohort analysis | Rate of customer/revenue loss; early warning on retention | Retention programs, product-market fit checks, CS focus | Actionable indicator; directly impacts LTV and growth trajectory |
Feature Adoption & Product Engagement | High, requires event tracking and behavioral analysis | Product instrumentation, analytics platform, segmentation | Depth of usage, feature value signals, churn predictors | Product roadmap, onboarding optimization, identifying expansion levers | Predicts retention and expansion; guides product investment |
Free-to-Paid Conversion Rate | Low–Medium, funnel tracking and cohort analysis | Signup/trial events, billing data, traffic attribution | Efficiency of product-led conversion from free to paid | Freemium/trial models, onboarding experiments, growth optimization | Lowers CAC; provides direct feedback on product-market fit |
Customer Health Score & Expansion Potential | High, composite model and ongoing validation | Integrated usage, support, NPS, revenue signals, modeling | Predictive churn risk and expansion opportunities; prioritization | Customer success prioritization, enterprise account management | Enables proactive retention and targeted expansion outreach |
North Star Metrics by Business Model
The eight metrics above are all legitimate north stars — but not all of them are the right starting point for every B2B SaaS business. Which one you prioritise depends on how your product creates and captures value. The biggest mistake founders make is picking a metric because it sounds sophisticated rather than because it maps to their actual GTM motion.
Business model | Primary north star | Supporting metrics | Why |
|---|---|---|---|
Product-led growth (PLG) | Free-to-Paid Conversion Rate | Feature adoption rate, time-to-value, activation rate | Revenue follows product value. Conversion rate tells you whether the product is delivering that value fast enough for users to pay. |
Sales-led growth (SLG) | Net Revenue Retention (NRR) | CAC payback period, MRR growth, expansion revenue | Sales cycles are expensive. NRR tells you whether closed accounts are growing — which determines whether the CAC investment compounds or leaks. |
Usage-based pricing | Feature Adoption + Expansion MRR | Usage growth per account, NRR, power-user depth | Revenue scales with usage. Tracking which features drive the most usage growth tells you where to invest product resources for the highest revenue return. |
Vertical SaaS | Customer Health Score | Churn rate, NPS, support ticket volume, feature breadth per account | Vertical SaaS wins through deep retention, not broad acquisition. Health score catches at-risk accounts before they churn — which matters more in a defined TAM where every customer is hard to replace. |
Infrastructure / API-first | Monthly Active Integrations / API Call Growth | CAC, expansion MRR, NRR | Stickiness comes from being embedded in customer workflows. Measuring active integration depth tells you how embedded you are — and how hard you'd be to rip out. |
Marketplace / network-effect SaaS | Gross Merchandise Value (GMV) or liquidity rate | Supply/demand balance, take rate, repeat transaction rate | Network-effect businesses live and die by liquidity — the percentage of supply that finds demand. Standard SaaS metrics undercount the real health signal. |
North Star Metrics by Company Stage
Your north star should also evolve as your company scales. A metric that's actionable at $50K ARR may be meaningless at $5M ARR — and vice versa. Here's a framework for which metrics deserve the most attention at each stage:
Stage | ARR range | Primary focus | Key north star | What to ignore |
|---|---|---|---|---|
Pre-PMF | $0–$500K | Does the product create enough value that users come back? | Feature adoption rate + Free-to-Paid conversion | NRR (not enough cohort history), CAC (deal volume too low to be statistically meaningful) |
Early growth | $500K–$3M | Is the unit economics story valid at small scale? | CAC payback period + MRR growth rate | LTV models (too early), complex health scores (not enough data) |
Scaling | $3M–$15M | Are we retaining and expanding revenue efficiently? | NRR + churn rate (segmented by cohort and plan) | Raw MRR growth (without NRR context it's misleading), vanity engagement metrics |
Growth | $15M–$50M | What's the ceiling on our best customer segment? | LTV:CAC ratio + expansion MRR by segment | Aggregate churn (use cohort churn instead), simple conversion rate |
Late-stage / pre-IPO | $50M+ | Is growth efficient and durable? | Rule of 40 (growth rate + profit margin), NRR, gross margin | Feature adoption (should be optimised by now), conversion rate (not the bottleneck) |
The trap most founders fall into is tracking too many metrics simultaneously — a dashboard with 15 KPIs is just noise with better formatting. Pick one primary north star per stage, track two or three supporting metrics, and make the rest available for deep dives when something looks off.
If you're building this tracking infrastructure manually in spreadsheets, you're adding lag at every layer. Statspresso's AI Data Chat connects directly to your SaaS data sources — ask "what's our NRR by cohort for accounts signed in Q3?" and get the answer in seconds, not a ticket queue.
From Metrics to Movement Find and Track Your North Star
Choosing a North Star Metric isn't a branding exercise. It's a decision about what kind of company you're building and what evidence you trust most. The right metric gives your team focus. The wrong one gives you alignment theater.
The pattern is usually straightforward. Early-stage SaaS companies should lean toward activation, engagement, or feature adoption. Those metrics tell you whether users are finding value at all. Growth-stage companies can move toward MRR, NRR, churn discipline, and LTV because they have enough customer history to support those views.
That said, no serious operator should worship one metric in isolation. MRR without churn context is misleading. Conversion without activation depth is fragile. Engagement without retention is vanity with better branding.
Stop Waiting for Data. Find Your North Star Now.
Your gut says growth is slowing, but the numbers are stuck behind a queue. You file a BI ticket, wait, get a dashboard later, and it still doesn't answer the follow-up question you care about. Waiting on data is a business killer. In 2026, you shouldn't need SQL, a sprint cycle, or a patient analyst just to understand what's happening in your own SaaS business.
That's why a North Star Metric matters. It gives the team one number that reflects real customer value and business progress. In B2B SaaS, the idea became mainstream when Amplitude pushed the concept into the market in 2016, and Finmark's compilation notes that more than 80 companies adopted specific North Star Metrics after that wave of thinking took hold, as summarized in CXL's review of North Star Metrics.
The hard part isn't understanding the concept. It's choosing the right metric for your model, stage, and motion. A PLG company shouldn't blindly copy a sales-led CRM. An AI analytics product shouldn't obsess over logins if the actual value is insight generation. Good North Stars focus teams. Bad ones create fancy dashboards and bad decisions.
This B2B SaaS North Star Metrics List gets to the point. You'll see which metrics fit which business model, where they work, where they mislead, and how to track them with Statspresso, a Conversational AI Data Analyst that lets you skip the SQL and just ask your data a question.
1. Monthly Recurring Revenue
A founder checks the board deck and sees MRR up. Good news, until the next question lands. Is growth coming from new customers, expansions, annual contracts normalized monthly, or a pricing change that won't repeat? If MRR is going to be your North Star, it has to explain the business, not just decorate the slide.
MRR earns its place because it ties product, sales, pricing, and retention into one operating number. For sales-led SaaS, that often makes it the right company-level metric once the team has a repeatable motion. For PLG SaaS, MRR is usually better as the top-line outcome, with activation, adoption, or free-to-paid conversion acting as the leading indicators underneath. Early-stage teams get into trouble when they force MRR to answer a product-market-fit question it cannot answer.
That trade-off matters. A mature sales-led company can run the business off MRR slices by segment, rep, and expansion path. A PLG company still needs to know what user behavior creates revenue later. If self-serve users never reach the core value moment, MRR will lag the problem by weeks or months.
When MRR works and when it doesn't
MRR works best when pricing is subscription-based, reporting is clean, and the team already knows the funnel stages that produce durable revenue.
Use it well:
Segment MRR: Break it out by plan, industry, sales-led vs. self-serve, or integration type.
Split new, expansion, contraction, and reactivation MRR: A single growth number hides too much.
Read it with product metrics: Pair MRR with activation, churn, and feature adoption so teams know what to fix.
Common mistakes:
Using one headline number for every decision: Finance can live with that. Operators cannot.
Treating all revenue as equal: Enterprise expansion and SMB new business behave differently.
Reviewing it only monthly: Weekly checkpoints catch changes in pipeline conversion and expansion earlier.
Practical rule: MRR is a strong scoreboard. It still needs leading indicators if you want to steer.
For teams using Statspresso, this metric gets more useful once you ask follow-up questions in plain English instead of waiting for a custom dashboard. Ask for MRR by acquisition motion, by workspace size, or by first integration connected. That is where a conversational AI data analyst helps. Founders and PMs can inspect revenue patterns day to day without handing every question to a data team.
A good setup starts with clean definitions. Decide whether discounts, services, annual prepaids, and paused accounts belong in reported MRR. If the team uses different logic across finance, product, and GTM, the metric becomes a source of arguments instead of decisions. For a useful primer, Statspresso's guide to understanding key performance indicators pairs well with this section, and this breakdown of calculating MRR for SaaS founders is a solid reference for the finance side.
Try asking Statspresso: “Show monthly recurring revenue for the last 12 months, split into new, expansion, contraction, and churn MRR, and compare self-serve vs. sales-led customers.”
2. Customer Acquisition Cost
A familiar SaaS problem: revenue is growing, the pipeline looks full, and the board still asks why cash burn is not improving. CAC usually explains the gap. If winning each new customer costs too much, growth adds pressure instead of margin.
Customer Acquisition Cost is the clearest test of whether your go-to-market motion scales. In an early search phase, CAC jumps around because the team is trying channels, pricing, and messaging. Once a motion starts repeating, CAC becomes a discipline metric. It shows whether demand gen, sales, onboarding, and product handoff work as one system or as four separate teams creating friction for each other.
CAC also changes meaning by business model.
For PLG SaaS, I care less about one blended number and more about CAC by channel and activation path. Paid social may generate signups. Organic search may bring in buyers with a specific reporting problem and a faster path to value. If the product gets users to a useful first output quickly, paid acquisition can work. If activation is weak, even cheap traffic becomes expensive.
For sales-led SaaS, segment-level CAC matters more than top-line averages. SMB, mid-market, and enterprise deals carry different sales cycles, implementation costs, and rep involvement. A blended CAC can hide a bad enterprise motion or make an efficient SMB engine look worse than it is. Hybrid companies need even more honesty. A “self-serve” funnel that depends on SDR rescue calls is not self-serve in cost terms.
That is why PMs should care about CAC, not just finance and growth leads. Friction in signup, weak onboarding, slow time to first value, and poor product-qualified lead signals all raise acquisition cost. CAC is partly a marketing number and partly a product operations number.
A few patterns are worth tracking closely:
PLG teams: Measure CAC by source, signup cohort, and activation milestone.
Sales-led teams: Measure CAC by segment, sales team, and payback period.
Hybrid teams: Separate true self-serve CAC from sales-assisted CAC.
Growth-stage teams: Watch trend lines weekly, not just monthly, because channel efficiency can deteriorate fast.
Cheap CAC from bad-fit customers still produces expensive growth later.
I rarely recommend CAC as the single North Star metric. It works better as a guardrail around growth because teams can improve CAC the wrong way by cutting spend too aggressively, narrowing ICP too far, or underinvesting in sales capacity. The better question is whether CAC is justified by retention, expansion, and payback for the customer type you are acquiring.
For day-to-day decision-making, a conversational AI analyst is often more useful than another static dashboard. In Statspresso, founders and PMs can ask for CAC by channel, by segment, by sales-assisted vs. self-serve, or by the first data source connected after signup. That last cut is especially useful in analytics products because it links acquisition cost to buyer intent, not just lead source. If renewal economics are part of the picture, teams can pair CAC analysis with the WhatPulse sales renewal tool to pressure-test whether acquisition costs still make sense after renewal and expansion assumptions.
Try asking Statspresso: “Show CAC for the last two quarters by acquisition channel, split self-serve vs. sales-assisted, and compare activated customers vs. non-activated signups.”
3. Net Revenue Retention
A company can post strong new bookings and still have a retention problem. Net Revenue Retention shows whether the installed base is getting more valuable over time.
That makes NRR one of the clearest strategic metrics in B2B SaaS, especially once the business has enough cohort history to trust the signal. It answers a tougher question than topline growth. Are existing customers renewing, expanding, and finding more reasons to stay?
For sales-led SaaS, NRR often reflects onboarding quality, account management, pricing design, and how well expansion paths match the buyer's rollout motion. For PLG companies, it usually rises when more teams activate, invite coworkers, connect more data, or adopt higher-usage workflows. Same metric, different engine.
What NRR reveals that MRR hides
MRR can increase while account quality gets worse. A busy sales team can mask poor retention for a while. NRR removes that cover because it isolates what happens after the initial sale.
Use NRR when:
You have real expansion mechanics: More seats, usage-based pricing, add-ons, extra modules, or multi-team rollout.
Customer value compounds over time: The product gets harder to replace after adoption spreads.
Product, success, and sales all shape outcomes: NRR captures the handoff between promise and delivered value.
NRR is less useful as the primary North Star when:
You are still very early: Small cohorts can swing wildly and create false confidence.
Pricing is mostly flat: If accounts have little room to grow, NRR behaves more like a churn metric.
Contracts are short and transactional: Retention noise can drown out the operating signal.
The practical move is to segment it. SMB, mid-market, enterprise, self-serve, and sales-assisted accounts each expand for different reasons. If one segment has high logo retention but weak expansion, that usually points to packaging or adoption gaps. If another expands well but churns after the first renewal, the problem is often implementation quality or a bad-fit ICP.
I also would not read NRR as a finance-only metric. It is one of the fastest ways to see whether your product's value moment turns into account growth. Teams that improve NRR usually do it by shortening time-to-value, increasing feature adoption in the first 60 to 90 days, and giving customers a clear reason to broaden usage inside the account.
That is why conversational analysis helps here. In Statspresso, a founder, PM, or CS leader can ask for NRR by cohort, segment, contract type, plan, owner, or activation milestone without waiting on a custom SQL pull. That matters most for companies without a dedicated data team, because the useful questions change every week. Which accounts expanded after adopting a specific feature? Which onboarding path correlates with stronger renewal? Which segment has healthy gross retention but weak net retention?
If your team is trying to improve renewals and expansion at the same time, the WhatPulse sales renewal tool is a practical way to model the commercial side alongside your NRR analysis.
Try asking Statspresso: “Show quarterly NRR for the last 8 quarters by segment and sales motion. Then identify the product actions shared by accounts that expanded in the last quarter.”
4. Customer Lifetime Value
A founder looks at CAC and feels good. The deals are closing, signups are up, and the pipeline looks healthy. Six months later, the same company realizes the new customers either never expanded, needed too much support, or left before payback. That is the problem LTV is supposed to catch.
Customer Lifetime Value measures how much economic value a customer produces across the full relationship. In B2B SaaS, that makes it a strategy metric, not just a finance metric. It helps answer a harder question than “Can we acquire customers?” It answers “Which customers are worth acquiring, keeping, and expanding?”
LTV gets more useful as the business matures. Early-stage teams usually do not have enough retention history to model it with confidence, especially in sales-led SaaS with annual contracts. In PLG, you can often read the signal sooner because activation, expansion, and early churn show up faster. Even then, I would treat early LTV as directional until the cohorts have aged.
A healthy LTV:CAC relationship matters, but the exact ratio depends on sales motion, gross margin, contract length, and cost to serve. A self-serve PLG product can tolerate a different profile than an enterprise product with long onboarding and heavy support. The practical rule is simple. If a segment looks expensive to acquire and slow to expand, its LTV needs to justify that burden.
The teams that use LTV well rarely look at one blended number. They break it apart so it can guide decisions:
By acquisition source: Organic, partner, outbound, and paid channels often produce very different customer durability.
By sales motion: PLG users who activate on their own behave differently from sales-assisted accounts.
By plan or contract type: Monthly self-serve customers and annual enterprise customers should not sit in the same LTV bucket.
By use case: Accounts that embed the product into a recurring workflow usually stay longer and expand more.
For a product like Statspresso, I would pay close attention to accounts that connect multiple data sources, return weekly, and share answers across functions. Those behaviors usually indicate the product moved from occasional reporting help to an operating habit. That is the kind of pattern that raises LTV in a durable way.
There are trade-offs here. Revenue-based LTV is fast to calculate, but it can mislead if onboarding, support, and success costs vary a lot by segment. Fully loaded LTV is more useful for decision-making, but it takes more discipline to maintain. The wrong move is using a single company-wide average and calling it strategy. That number hides both your best-fit customers and the segments draining time.
Statspresso is useful here because LTV questions change constantly, and smaller teams rarely have an analyst available for every follow-up. A founder can ask which 2024 cohorts have the highest projected LTV, whether sales-led accounts outperform PLG after 12 months, or which activation events correlate with longer retention. That lets teams investigate day-to-day without building a queue of SQL requests.
LTV is a check on acquisition quality, retention quality, and expansion quality at the same time.
Try asking Statspresso: “Compare projected LTV by acquisition source, plan, and sales motion for customers acquired in 2024 and 2025. Then show which activation behaviors are most common in the highest-LTV cohorts.”
5. Churn Rate Customer and Revenue
Monday morning looks great until the renewal report lands. New bookings are up, trial volume is healthy, and the product team is celebrating engagement. Then three good accounts cancel and two larger customers downgrade. Growth did not disappear. It was offset.
That is why churn belongs on any serious North Star shortlist. It tests whether growth is durable or just loud. I look at customer churn and revenue churn together because they answer different operating questions. Customer churn shows logo loss. Revenue churn shows whether the accounts leaving matter to the business.
As noted earlier, strong SaaS companies keep a tight grip on churn. The useful takeaway is not a universal benchmark. It is the operating habit behind it. Teams that review churn by segment, cohort, and product behavior catch weak retention earlier and make better decisions about acquisition, onboarding, and pricing.
Customer churn and revenue churn answer different questions
A PLG company can tolerate more logo churn if low-value self-serve users leave while strong-fit accounts activate, adopt, and expand. A sales-led company has the opposite risk. Logo counts can look stable while revenue churn creeps up because a handful of larger customers cut seats, downgrade plans, or fail to renew.
This is why a single churn number is not enough.
Track churn at least three ways:
By plan: Self-serve, mid-market, and enterprise tiers rarely churn for the same reason.
By company size: SMB churn often reflects weak onboarding or low urgency. Enterprise churn often points to rollout failure, missing governance, or poor executive buy-in.
By activation or usage depth: Teams that never hit the core value moment should not be mixed in with accounts that have reached full adoption.
Cohorts matter just as much. If customers acquired in one quarter churn faster than the rest, I would first inspect channel mix, onboarding changes, pricing updates, and product releases from that period. Without cohort analysis, churn stays abstract. With cohorts, it becomes diagnosable.
The trade-off is straightforward. Customer churn is easier to explain to the whole company. Revenue churn is better for prioritization because it tells you where money is leaking. Early-stage PLG teams often start with logo churn because volume is higher and contracts are smaller. Later-stage and sales-led teams should put more weight on revenue churn, gross and net, because one account can move the entire month.
Statspresso helps because churn questions rarely stop at one chart. A founder or PM can ask which cohorts have the worst six-month revenue churn, whether sales-assisted accounts retain better than pure PLG signups, or whether teams that connect a second data source are less likely to cancel. That kind of day-to-day analysis usually gets stuck in a BI queue. It should not.
Try asking Statspresso: “Show monthly customer churn and revenue churn for the past 12 months, broken out by plan and acquisition source. Then compare churn for accounts that reached activation in their first 30 days versus those that did not.”
6. Feature Adoption and Product Engagement
Many SaaS teams should start here, even if they don't. If you haven't nailed what real product value looks like, feature adoption and engagement are often better North Star candidates than revenue.
A North Star should reflect customer value, not just attention. That's why generic logins and pageviews are weak choices. The right engagement metric captures the moment the product becomes useful.
A strong B2B SaaS example comes from Dropbox. Finmark's compilation notes that Dropbox used “trial accounts with more than 3 active users in week 1,” and that metric predicted conversion to paid subscriptions at 40% to 60% based on internal benchmarks shared in Amplitude case studies, as summarized in CXL's North Star analysis. That's a great North Star because it captures collaborative value early, not passive traffic.
Good engagement metrics are specific
For analytics tools, “active users” is often too vague. A user can open the app, poke around, and leave without getting value. A better metric is tied to the useful job.
For Statspresso, better candidates include:
AI queries resolved per active workspace
Insights added to dashboards
Shared dashboards viewed by teams
Connected sources actively queried
The verified brief explicitly points to a gap in the market. Existing North Star content covers generic SaaS examples but rarely gives AI-specific guidance for products where value comes from generated insights instead of simple activity, as described in Growth Academy's North Star examples article. That gap is real. AI products need value metrics, not vanity traffic metrics.
What works here is choosing one action that means the user got something useful. What doesn't work is bundling every product event into an “engagement score” nobody trusts.
If your engagement metric goes up while retention stays weak, the metric is too shallow.
Try asking Statspresso: “What percentage of active users have adopted the AI Insight Gallery feature? Show me the trend over the last 6 months.”
7. Free-to-Paid Conversion Rate
A familiar PLG problem. Signups look healthy, the top of funnel keeps growing, and revenue still feels stuck.
Free-to-paid conversion cuts through that noise. It shows whether the product gets users to a buying moment, not just an account-creation moment.
This metric matters most for freemium, free trial, and hybrid SaaS models where users can experience value before talking to sales. In a sales-led company, it usually belongs lower in the dashboard. Procurement, security review, and contract timing can delay purchase long after the product has proved its value, so conversion inside the app is not always the cleanest company-level North Star.
What improves conversion is activation. Users pay after they reach a clear value milestone and understand why the product belongs in their workflow. As noted earlier, teams that anchor on first-value behavior usually make better decisions than teams staring at raw activity counts.
For Statspresso, I would track trial users who:
connect a real data source
ask several real business questions
save or share an output with a teammate
come back after the first session
Those actions show buying intent far better than logins or time in app. A user can spend twenty minutes clicking around and still be nowhere near a purchase decision.
There is a trade-off here. Push upgrade prompts too early and conversion can rise for a month while retention gets worse a quarter later. Wait too long and qualified users stay on free plans that never turn into revenue. The right threshold depends on the model. PLG products usually convert best right after a user gets a concrete result. Sales-led products often use product activity as a qualification signal for sales, not the final conversion event itself.
Use free-to-paid conversion as a North Star if the product does the selling. Treat it as a supporting metric if sales, legal review, or multi-stakeholder buying committees do the selling.
Try asking Statspresso: “What's our trial-to-paid conversion rate for cohorts who signed up in the last 3 months? Break it down by traffic source, and compare users who connected a data source in week one vs. those who didn't.”
8. Customer Health Score and Expansion Revenue Potential
A customer looks fine at renewal prep. Usage is still there. The champion still shows up to calls. Then procurement pushes the deal into a downsize, or a competitor gets invited into the account. That usually happens because the team tracked activity, but missed account health.
Customer health score is a working model for retention and expansion. It pulls together the signals that matter across product, success, support, and commercial teams. Used well, it answers two practical questions: which accounts need intervention now, and which accounts have earned an upsell conversation.
This metric is especially useful once a SaaS business has enough customers that account reviews become inconsistent. Earlier-stage PLG companies often use lightweight health logic to spot accounts that are ready for sales assist. Sales-led teams usually need a richer score because renewals, stakeholder adoption, and rollout depth matter more than raw logins.
The trade-off is clarity versus accuracy. Add too many inputs and the score turns into a black box that nobody trusts. Keep it too simple and it misses real risk. The best version is explainable in one minute by a CSM, PM, or founder.
For Statspresso, I'd build the score from account-level signals such as:
question volume over the last 30 days
number of active users per workspace
dashboards created, saved, or shared
connector setup depth across real business data sources
support tickets by type and resolution pattern
executive or cross-functional usage, not just one analyst
contract tier, renewal date, and recent seat changes
Then map the score to action. High-health accounts with broad usage and growing team adoption go to expansion review. Mid-health accounts get targeted enablement. Low-health accounts need root-cause analysis, usually around setup gaps, weak rollout, or low value realization.
Expansion potential should not be treated as the same thing as health. Some accounts are healthy but fully saturated on their current plan. Others have uneven health and still hold upside if one department is getting strong value while the rest of the company has not adopted yet. That distinction matters more in sales-led SaaS, where expansion often depends on stakeholder coverage and procurement timing, not just product engagement.
For founders without a data team, this is exactly the kind of metric that should be easy to query in plain English instead of buried in a CRM plus BI stack. Statspresso can pull product usage, billing context, and support signals into one view so teams can ask follow-up questions on the fly, without waiting on an analyst to rebuild a dashboard.
A useful health score changes who gets attention this week and why.
Try asking Statspresso: “Show me customers with high product engagement, low support friction, renewal in the next 120 days, and unused seat or workspace expansion potential. Group them by plan and flag the best upsell candidates.”
B2B SaaS: 8 North Star Metrics Comparison
Metric | Implementation complexity | Resource requirements | Expected outcomes | Ideal use cases | Key advantages |
|---|---|---|---|---|---|
Monthly Recurring Revenue (MRR) | Low, simple subscription math | Billing & subscription data, basic dashboards | Predictable monthly revenue and growth trends | Regular revenue monitoring, forecasting, investor reporting | Easy to communicate, central for valuation and forecasts |
Customer Acquisition Cost (CAC) | Medium, cost allocation & attribution required | Marketing & sales spend, CRM, attribution tooling | Cost per new customer and channel efficiency insights | Optimizing acquisition spend and go-to-market strategy | Reveals channel ROI; informs budget and payback analysis |
Net Revenue Retention (NRR) | High, track expansions, contractions & churn by cohort | Detailed MRR movements, billing changes, cohort analytics | Revenue retention including expansion; indicates sustainability (>100%) | Mature SaaS growth assessment, upsell effectiveness, investor evaluation | Shows expansion power; reduces reliance on new acquisition |
Customer Lifetime Value (LTV) | Medium, needs churn, ARPU and margin modeling | ARPU, churn rates, gross margin, historical cohorts | Estimated long-term revenue/profit per customer | Setting CAC targets, pricing strategy, segment prioritization | Guides acquisition spend; highlights high-value segments |
Churn Rate (Customer & Revenue) | Low–Medium, straightforward calc, needs segmentation | Customer records, revenue tracking, cohort analysis | Rate of customer/revenue loss; early warning on retention | Retention programs, product-market fit checks, CS focus | Actionable indicator; directly impacts LTV and growth trajectory |
Feature Adoption & Product Engagement | High, requires event tracking and behavioral analysis | Product instrumentation, analytics platform, segmentation | Depth of usage, feature value signals, churn predictors | Product roadmap, onboarding optimization, identifying expansion levers | Predicts retention and expansion; guides product investment |
Free-to-Paid Conversion Rate | Low–Medium, funnel tracking and cohort analysis | Signup/trial events, billing data, traffic attribution | Efficiency of product-led conversion from free to paid | Freemium/trial models, onboarding experiments, growth optimization | Lowers CAC; provides direct feedback on product-market fit |
Customer Health Score & Expansion Potential | High, composite model and ongoing validation | Integrated usage, support, NPS, revenue signals, modeling | Predictive churn risk and expansion opportunities; prioritization | Customer success prioritization, enterprise account management | Enables proactive retention and targeted expansion outreach |
North Star Metrics by Business Model
The eight metrics above are all legitimate north stars — but not all of them are the right starting point for every B2B SaaS business. Which one you prioritise depends on how your product creates and captures value. The biggest mistake founders make is picking a metric because it sounds sophisticated rather than because it maps to their actual GTM motion.
Business model | Primary north star | Supporting metrics | Why |
|---|---|---|---|
Product-led growth (PLG) | Free-to-Paid Conversion Rate | Feature adoption rate, time-to-value, activation rate | Revenue follows product value. Conversion rate tells you whether the product is delivering that value fast enough for users to pay. |
Sales-led growth (SLG) | Net Revenue Retention (NRR) | CAC payback period, MRR growth, expansion revenue | Sales cycles are expensive. NRR tells you whether closed accounts are growing — which determines whether the CAC investment compounds or leaks. |
Usage-based pricing | Feature Adoption + Expansion MRR | Usage growth per account, NRR, power-user depth | Revenue scales with usage. Tracking which features drive the most usage growth tells you where to invest product resources for the highest revenue return. |
Vertical SaaS | Customer Health Score | Churn rate, NPS, support ticket volume, feature breadth per account | Vertical SaaS wins through deep retention, not broad acquisition. Health score catches at-risk accounts before they churn — which matters more in a defined TAM where every customer is hard to replace. |
Infrastructure / API-first | Monthly Active Integrations / API Call Growth | CAC, expansion MRR, NRR | Stickiness comes from being embedded in customer workflows. Measuring active integration depth tells you how embedded you are — and how hard you'd be to rip out. |
Marketplace / network-effect SaaS | Gross Merchandise Value (GMV) or liquidity rate | Supply/demand balance, take rate, repeat transaction rate | Network-effect businesses live and die by liquidity — the percentage of supply that finds demand. Standard SaaS metrics undercount the real health signal. |
North Star Metrics by Company Stage
Your north star should also evolve as your company scales. A metric that's actionable at $50K ARR may be meaningless at $5M ARR — and vice versa. Here's a framework for which metrics deserve the most attention at each stage:
Stage | ARR range | Primary focus | Key north star | What to ignore |
|---|---|---|---|---|
Pre-PMF | $0–$500K | Does the product create enough value that users come back? | Feature adoption rate + Free-to-Paid conversion | NRR (not enough cohort history), CAC (deal volume too low to be statistically meaningful) |
Early growth | $500K–$3M | Is the unit economics story valid at small scale? | CAC payback period + MRR growth rate | LTV models (too early), complex health scores (not enough data) |
Scaling | $3M–$15M | Are we retaining and expanding revenue efficiently? | NRR + churn rate (segmented by cohort and plan) | Raw MRR growth (without NRR context it's misleading), vanity engagement metrics |
Growth | $15M–$50M | What's the ceiling on our best customer segment? | LTV:CAC ratio + expansion MRR by segment | Aggregate churn (use cohort churn instead), simple conversion rate |
Late-stage / pre-IPO | $50M+ | Is growth efficient and durable? | Rule of 40 (growth rate + profit margin), NRR, gross margin | Feature adoption (should be optimised by now), conversion rate (not the bottleneck) |
The trap most founders fall into is tracking too many metrics simultaneously — a dashboard with 15 KPIs is just noise with better formatting. Pick one primary north star per stage, track two or three supporting metrics, and make the rest available for deep dives when something looks off.
If you're building this tracking infrastructure manually in spreadsheets, you're adding lag at every layer. Statspresso's AI Data Chat connects directly to your SaaS data sources — ask "what's our NRR by cohort for accounts signed in Q3?" and get the answer in seconds, not a ticket queue.
From Metrics to Movement Find and Track Your North Star
Choosing a North Star Metric isn't a branding exercise. It's a decision about what kind of company you're building and what evidence you trust most. The right metric gives your team focus. The wrong one gives you alignment theater.
The pattern is usually straightforward. Early-stage SaaS companies should lean toward activation, engagement, or feature adoption. Those metrics tell you whether users are finding value at all. Growth-stage companies can move toward MRR, NRR, churn discipline, and LTV because they have enough customer history to support those views.
That said, no serious operator should worship one metric in isolation. MRR without churn context is misleading. Conversion without activation depth is fragile. Engagement without retention is vanity with better branding.
Stop Waiting for Data. Find Your North Star Now.
Your gut says growth is slowing, but the numbers are stuck behind a queue. You file a BI ticket, wait, get a dashboard later, and it still doesn't answer the follow-up question you care about. Waiting on data is a business killer. In 2026, you shouldn't need SQL, a sprint cycle, or a patient analyst just to understand what's happening in your own SaaS business.
That's why a North Star Metric matters. It gives the team one number that reflects real customer value and business progress. In B2B SaaS, the idea became mainstream when Amplitude pushed the concept into the market in 2016, and Finmark's compilation notes that more than 80 companies adopted specific North Star Metrics after that wave of thinking took hold, as summarized in CXL's review of North Star Metrics.
The hard part isn't understanding the concept. It's choosing the right metric for your model, stage, and motion. A PLG company shouldn't blindly copy a sales-led CRM. An AI analytics product shouldn't obsess over logins if the actual value is insight generation. Good North Stars focus teams. Bad ones create fancy dashboards and bad decisions.
This B2B SaaS North Star Metrics List gets to the point. You'll see which metrics fit which business model, where they work, where they mislead, and how to track them with Statspresso, a Conversational AI Data Analyst that lets you skip the SQL and just ask your data a question.
1. Monthly Recurring Revenue
A founder checks the board deck and sees MRR up. Good news, until the next question lands. Is growth coming from new customers, expansions, annual contracts normalized monthly, or a pricing change that won't repeat? If MRR is going to be your North Star, it has to explain the business, not just decorate the slide.
MRR earns its place because it ties product, sales, pricing, and retention into one operating number. For sales-led SaaS, that often makes it the right company-level metric once the team has a repeatable motion. For PLG SaaS, MRR is usually better as the top-line outcome, with activation, adoption, or free-to-paid conversion acting as the leading indicators underneath. Early-stage teams get into trouble when they force MRR to answer a product-market-fit question it cannot answer.
That trade-off matters. A mature sales-led company can run the business off MRR slices by segment, rep, and expansion path. A PLG company still needs to know what user behavior creates revenue later. If self-serve users never reach the core value moment, MRR will lag the problem by weeks or months.
When MRR works and when it doesn't
MRR works best when pricing is subscription-based, reporting is clean, and the team already knows the funnel stages that produce durable revenue.
Use it well:
Segment MRR: Break it out by plan, industry, sales-led vs. self-serve, or integration type.
Split new, expansion, contraction, and reactivation MRR: A single growth number hides too much.
Read it with product metrics: Pair MRR with activation, churn, and feature adoption so teams know what to fix.
Common mistakes:
Using one headline number for every decision: Finance can live with that. Operators cannot.
Treating all revenue as equal: Enterprise expansion and SMB new business behave differently.
Reviewing it only monthly: Weekly checkpoints catch changes in pipeline conversion and expansion earlier.
Practical rule: MRR is a strong scoreboard. It still needs leading indicators if you want to steer.
For teams using Statspresso, this metric gets more useful once you ask follow-up questions in plain English instead of waiting for a custom dashboard. Ask for MRR by acquisition motion, by workspace size, or by first integration connected. That is where a conversational AI data analyst helps. Founders and PMs can inspect revenue patterns day to day without handing every question to a data team.
A good setup starts with clean definitions. Decide whether discounts, services, annual prepaids, and paused accounts belong in reported MRR. If the team uses different logic across finance, product, and GTM, the metric becomes a source of arguments instead of decisions. For a useful primer, Statspresso's guide to understanding key performance indicators pairs well with this section, and this breakdown of calculating MRR for SaaS founders is a solid reference for the finance side.
Try asking Statspresso: “Show monthly recurring revenue for the last 12 months, split into new, expansion, contraction, and churn MRR, and compare self-serve vs. sales-led customers.”
2. Customer Acquisition Cost
A familiar SaaS problem: revenue is growing, the pipeline looks full, and the board still asks why cash burn is not improving. CAC usually explains the gap. If winning each new customer costs too much, growth adds pressure instead of margin.
Customer Acquisition Cost is the clearest test of whether your go-to-market motion scales. In an early search phase, CAC jumps around because the team is trying channels, pricing, and messaging. Once a motion starts repeating, CAC becomes a discipline metric. It shows whether demand gen, sales, onboarding, and product handoff work as one system or as four separate teams creating friction for each other.
CAC also changes meaning by business model.
For PLG SaaS, I care less about one blended number and more about CAC by channel and activation path. Paid social may generate signups. Organic search may bring in buyers with a specific reporting problem and a faster path to value. If the product gets users to a useful first output quickly, paid acquisition can work. If activation is weak, even cheap traffic becomes expensive.
For sales-led SaaS, segment-level CAC matters more than top-line averages. SMB, mid-market, and enterprise deals carry different sales cycles, implementation costs, and rep involvement. A blended CAC can hide a bad enterprise motion or make an efficient SMB engine look worse than it is. Hybrid companies need even more honesty. A “self-serve” funnel that depends on SDR rescue calls is not self-serve in cost terms.
That is why PMs should care about CAC, not just finance and growth leads. Friction in signup, weak onboarding, slow time to first value, and poor product-qualified lead signals all raise acquisition cost. CAC is partly a marketing number and partly a product operations number.
A few patterns are worth tracking closely:
PLG teams: Measure CAC by source, signup cohort, and activation milestone.
Sales-led teams: Measure CAC by segment, sales team, and payback period.
Hybrid teams: Separate true self-serve CAC from sales-assisted CAC.
Growth-stage teams: Watch trend lines weekly, not just monthly, because channel efficiency can deteriorate fast.
Cheap CAC from bad-fit customers still produces expensive growth later.
I rarely recommend CAC as the single North Star metric. It works better as a guardrail around growth because teams can improve CAC the wrong way by cutting spend too aggressively, narrowing ICP too far, or underinvesting in sales capacity. The better question is whether CAC is justified by retention, expansion, and payback for the customer type you are acquiring.
For day-to-day decision-making, a conversational AI analyst is often more useful than another static dashboard. In Statspresso, founders and PMs can ask for CAC by channel, by segment, by sales-assisted vs. self-serve, or by the first data source connected after signup. That last cut is especially useful in analytics products because it links acquisition cost to buyer intent, not just lead source. If renewal economics are part of the picture, teams can pair CAC analysis with the WhatPulse sales renewal tool to pressure-test whether acquisition costs still make sense after renewal and expansion assumptions.
Try asking Statspresso: “Show CAC for the last two quarters by acquisition channel, split self-serve vs. sales-assisted, and compare activated customers vs. non-activated signups.”
3. Net Revenue Retention
A company can post strong new bookings and still have a retention problem. Net Revenue Retention shows whether the installed base is getting more valuable over time.
That makes NRR one of the clearest strategic metrics in B2B SaaS, especially once the business has enough cohort history to trust the signal. It answers a tougher question than topline growth. Are existing customers renewing, expanding, and finding more reasons to stay?
For sales-led SaaS, NRR often reflects onboarding quality, account management, pricing design, and how well expansion paths match the buyer's rollout motion. For PLG companies, it usually rises when more teams activate, invite coworkers, connect more data, or adopt higher-usage workflows. Same metric, different engine.
What NRR reveals that MRR hides
MRR can increase while account quality gets worse. A busy sales team can mask poor retention for a while. NRR removes that cover because it isolates what happens after the initial sale.
Use NRR when:
You have real expansion mechanics: More seats, usage-based pricing, add-ons, extra modules, or multi-team rollout.
Customer value compounds over time: The product gets harder to replace after adoption spreads.
Product, success, and sales all shape outcomes: NRR captures the handoff between promise and delivered value.
NRR is less useful as the primary North Star when:
You are still very early: Small cohorts can swing wildly and create false confidence.
Pricing is mostly flat: If accounts have little room to grow, NRR behaves more like a churn metric.
Contracts are short and transactional: Retention noise can drown out the operating signal.
The practical move is to segment it. SMB, mid-market, enterprise, self-serve, and sales-assisted accounts each expand for different reasons. If one segment has high logo retention but weak expansion, that usually points to packaging or adoption gaps. If another expands well but churns after the first renewal, the problem is often implementation quality or a bad-fit ICP.
I also would not read NRR as a finance-only metric. It is one of the fastest ways to see whether your product's value moment turns into account growth. Teams that improve NRR usually do it by shortening time-to-value, increasing feature adoption in the first 60 to 90 days, and giving customers a clear reason to broaden usage inside the account.
That is why conversational analysis helps here. In Statspresso, a founder, PM, or CS leader can ask for NRR by cohort, segment, contract type, plan, owner, or activation milestone without waiting on a custom SQL pull. That matters most for companies without a dedicated data team, because the useful questions change every week. Which accounts expanded after adopting a specific feature? Which onboarding path correlates with stronger renewal? Which segment has healthy gross retention but weak net retention?
If your team is trying to improve renewals and expansion at the same time, the WhatPulse sales renewal tool is a practical way to model the commercial side alongside your NRR analysis.
Try asking Statspresso: “Show quarterly NRR for the last 8 quarters by segment and sales motion. Then identify the product actions shared by accounts that expanded in the last quarter.”
4. Customer Lifetime Value
A founder looks at CAC and feels good. The deals are closing, signups are up, and the pipeline looks healthy. Six months later, the same company realizes the new customers either never expanded, needed too much support, or left before payback. That is the problem LTV is supposed to catch.
Customer Lifetime Value measures how much economic value a customer produces across the full relationship. In B2B SaaS, that makes it a strategy metric, not just a finance metric. It helps answer a harder question than “Can we acquire customers?” It answers “Which customers are worth acquiring, keeping, and expanding?”
LTV gets more useful as the business matures. Early-stage teams usually do not have enough retention history to model it with confidence, especially in sales-led SaaS with annual contracts. In PLG, you can often read the signal sooner because activation, expansion, and early churn show up faster. Even then, I would treat early LTV as directional until the cohorts have aged.
A healthy LTV:CAC relationship matters, but the exact ratio depends on sales motion, gross margin, contract length, and cost to serve. A self-serve PLG product can tolerate a different profile than an enterprise product with long onboarding and heavy support. The practical rule is simple. If a segment looks expensive to acquire and slow to expand, its LTV needs to justify that burden.
The teams that use LTV well rarely look at one blended number. They break it apart so it can guide decisions:
By acquisition source: Organic, partner, outbound, and paid channels often produce very different customer durability.
By sales motion: PLG users who activate on their own behave differently from sales-assisted accounts.
By plan or contract type: Monthly self-serve customers and annual enterprise customers should not sit in the same LTV bucket.
By use case: Accounts that embed the product into a recurring workflow usually stay longer and expand more.
For a product like Statspresso, I would pay close attention to accounts that connect multiple data sources, return weekly, and share answers across functions. Those behaviors usually indicate the product moved from occasional reporting help to an operating habit. That is the kind of pattern that raises LTV in a durable way.
There are trade-offs here. Revenue-based LTV is fast to calculate, but it can mislead if onboarding, support, and success costs vary a lot by segment. Fully loaded LTV is more useful for decision-making, but it takes more discipline to maintain. The wrong move is using a single company-wide average and calling it strategy. That number hides both your best-fit customers and the segments draining time.
Statspresso is useful here because LTV questions change constantly, and smaller teams rarely have an analyst available for every follow-up. A founder can ask which 2024 cohorts have the highest projected LTV, whether sales-led accounts outperform PLG after 12 months, or which activation events correlate with longer retention. That lets teams investigate day-to-day without building a queue of SQL requests.
LTV is a check on acquisition quality, retention quality, and expansion quality at the same time.
Try asking Statspresso: “Compare projected LTV by acquisition source, plan, and sales motion for customers acquired in 2024 and 2025. Then show which activation behaviors are most common in the highest-LTV cohorts.”
5. Churn Rate Customer and Revenue
Monday morning looks great until the renewal report lands. New bookings are up, trial volume is healthy, and the product team is celebrating engagement. Then three good accounts cancel and two larger customers downgrade. Growth did not disappear. It was offset.
That is why churn belongs on any serious North Star shortlist. It tests whether growth is durable or just loud. I look at customer churn and revenue churn together because they answer different operating questions. Customer churn shows logo loss. Revenue churn shows whether the accounts leaving matter to the business.
As noted earlier, strong SaaS companies keep a tight grip on churn. The useful takeaway is not a universal benchmark. It is the operating habit behind it. Teams that review churn by segment, cohort, and product behavior catch weak retention earlier and make better decisions about acquisition, onboarding, and pricing.
Customer churn and revenue churn answer different questions
A PLG company can tolerate more logo churn if low-value self-serve users leave while strong-fit accounts activate, adopt, and expand. A sales-led company has the opposite risk. Logo counts can look stable while revenue churn creeps up because a handful of larger customers cut seats, downgrade plans, or fail to renew.
This is why a single churn number is not enough.
Track churn at least three ways:
By plan: Self-serve, mid-market, and enterprise tiers rarely churn for the same reason.
By company size: SMB churn often reflects weak onboarding or low urgency. Enterprise churn often points to rollout failure, missing governance, or poor executive buy-in.
By activation or usage depth: Teams that never hit the core value moment should not be mixed in with accounts that have reached full adoption.
Cohorts matter just as much. If customers acquired in one quarter churn faster than the rest, I would first inspect channel mix, onboarding changes, pricing updates, and product releases from that period. Without cohort analysis, churn stays abstract. With cohorts, it becomes diagnosable.
The trade-off is straightforward. Customer churn is easier to explain to the whole company. Revenue churn is better for prioritization because it tells you where money is leaking. Early-stage PLG teams often start with logo churn because volume is higher and contracts are smaller. Later-stage and sales-led teams should put more weight on revenue churn, gross and net, because one account can move the entire month.
Statspresso helps because churn questions rarely stop at one chart. A founder or PM can ask which cohorts have the worst six-month revenue churn, whether sales-assisted accounts retain better than pure PLG signups, or whether teams that connect a second data source are less likely to cancel. That kind of day-to-day analysis usually gets stuck in a BI queue. It should not.
Try asking Statspresso: “Show monthly customer churn and revenue churn for the past 12 months, broken out by plan and acquisition source. Then compare churn for accounts that reached activation in their first 30 days versus those that did not.”
6. Feature Adoption and Product Engagement
Many SaaS teams should start here, even if they don't. If you haven't nailed what real product value looks like, feature adoption and engagement are often better North Star candidates than revenue.
A North Star should reflect customer value, not just attention. That's why generic logins and pageviews are weak choices. The right engagement metric captures the moment the product becomes useful.
A strong B2B SaaS example comes from Dropbox. Finmark's compilation notes that Dropbox used “trial accounts with more than 3 active users in week 1,” and that metric predicted conversion to paid subscriptions at 40% to 60% based on internal benchmarks shared in Amplitude case studies, as summarized in CXL's North Star analysis. That's a great North Star because it captures collaborative value early, not passive traffic.
Good engagement metrics are specific
For analytics tools, “active users” is often too vague. A user can open the app, poke around, and leave without getting value. A better metric is tied to the useful job.
For Statspresso, better candidates include:
AI queries resolved per active workspace
Insights added to dashboards
Shared dashboards viewed by teams
Connected sources actively queried
The verified brief explicitly points to a gap in the market. Existing North Star content covers generic SaaS examples but rarely gives AI-specific guidance for products where value comes from generated insights instead of simple activity, as described in Growth Academy's North Star examples article. That gap is real. AI products need value metrics, not vanity traffic metrics.
What works here is choosing one action that means the user got something useful. What doesn't work is bundling every product event into an “engagement score” nobody trusts.
If your engagement metric goes up while retention stays weak, the metric is too shallow.
Try asking Statspresso: “What percentage of active users have adopted the AI Insight Gallery feature? Show me the trend over the last 6 months.”
7. Free-to-Paid Conversion Rate
A familiar PLG problem. Signups look healthy, the top of funnel keeps growing, and revenue still feels stuck.
Free-to-paid conversion cuts through that noise. It shows whether the product gets users to a buying moment, not just an account-creation moment.
This metric matters most for freemium, free trial, and hybrid SaaS models where users can experience value before talking to sales. In a sales-led company, it usually belongs lower in the dashboard. Procurement, security review, and contract timing can delay purchase long after the product has proved its value, so conversion inside the app is not always the cleanest company-level North Star.
What improves conversion is activation. Users pay after they reach a clear value milestone and understand why the product belongs in their workflow. As noted earlier, teams that anchor on first-value behavior usually make better decisions than teams staring at raw activity counts.
For Statspresso, I would track trial users who:
connect a real data source
ask several real business questions
save or share an output with a teammate
come back after the first session
Those actions show buying intent far better than logins or time in app. A user can spend twenty minutes clicking around and still be nowhere near a purchase decision.
There is a trade-off here. Push upgrade prompts too early and conversion can rise for a month while retention gets worse a quarter later. Wait too long and qualified users stay on free plans that never turn into revenue. The right threshold depends on the model. PLG products usually convert best right after a user gets a concrete result. Sales-led products often use product activity as a qualification signal for sales, not the final conversion event itself.
Use free-to-paid conversion as a North Star if the product does the selling. Treat it as a supporting metric if sales, legal review, or multi-stakeholder buying committees do the selling.
Try asking Statspresso: “What's our trial-to-paid conversion rate for cohorts who signed up in the last 3 months? Break it down by traffic source, and compare users who connected a data source in week one vs. those who didn't.”
8. Customer Health Score and Expansion Revenue Potential
A customer looks fine at renewal prep. Usage is still there. The champion still shows up to calls. Then procurement pushes the deal into a downsize, or a competitor gets invited into the account. That usually happens because the team tracked activity, but missed account health.
Customer health score is a working model for retention and expansion. It pulls together the signals that matter across product, success, support, and commercial teams. Used well, it answers two practical questions: which accounts need intervention now, and which accounts have earned an upsell conversation.
This metric is especially useful once a SaaS business has enough customers that account reviews become inconsistent. Earlier-stage PLG companies often use lightweight health logic to spot accounts that are ready for sales assist. Sales-led teams usually need a richer score because renewals, stakeholder adoption, and rollout depth matter more than raw logins.
The trade-off is clarity versus accuracy. Add too many inputs and the score turns into a black box that nobody trusts. Keep it too simple and it misses real risk. The best version is explainable in one minute by a CSM, PM, or founder.
For Statspresso, I'd build the score from account-level signals such as:
question volume over the last 30 days
number of active users per workspace
dashboards created, saved, or shared
connector setup depth across real business data sources
support tickets by type and resolution pattern
executive or cross-functional usage, not just one analyst
contract tier, renewal date, and recent seat changes
Then map the score to action. High-health accounts with broad usage and growing team adoption go to expansion review. Mid-health accounts get targeted enablement. Low-health accounts need root-cause analysis, usually around setup gaps, weak rollout, or low value realization.
Expansion potential should not be treated as the same thing as health. Some accounts are healthy but fully saturated on their current plan. Others have uneven health and still hold upside if one department is getting strong value while the rest of the company has not adopted yet. That distinction matters more in sales-led SaaS, where expansion often depends on stakeholder coverage and procurement timing, not just product engagement.
For founders without a data team, this is exactly the kind of metric that should be easy to query in plain English instead of buried in a CRM plus BI stack. Statspresso can pull product usage, billing context, and support signals into one view so teams can ask follow-up questions on the fly, without waiting on an analyst to rebuild a dashboard.
A useful health score changes who gets attention this week and why.
Try asking Statspresso: “Show me customers with high product engagement, low support friction, renewal in the next 120 days, and unused seat or workspace expansion potential. Group them by plan and flag the best upsell candidates.”
B2B SaaS: 8 North Star Metrics Comparison
Metric | Implementation complexity | Resource requirements | Expected outcomes | Ideal use cases | Key advantages |
|---|---|---|---|---|---|
Monthly Recurring Revenue (MRR) | Low, simple subscription math | Billing & subscription data, basic dashboards | Predictable monthly revenue and growth trends | Regular revenue monitoring, forecasting, investor reporting | Easy to communicate, central for valuation and forecasts |
Customer Acquisition Cost (CAC) | Medium, cost allocation & attribution required | Marketing & sales spend, CRM, attribution tooling | Cost per new customer and channel efficiency insights | Optimizing acquisition spend and go-to-market strategy | Reveals channel ROI; informs budget and payback analysis |
Net Revenue Retention (NRR) | High, track expansions, contractions & churn by cohort | Detailed MRR movements, billing changes, cohort analytics | Revenue retention including expansion; indicates sustainability (>100%) | Mature SaaS growth assessment, upsell effectiveness, investor evaluation | Shows expansion power; reduces reliance on new acquisition |
Customer Lifetime Value (LTV) | Medium, needs churn, ARPU and margin modeling | ARPU, churn rates, gross margin, historical cohorts | Estimated long-term revenue/profit per customer | Setting CAC targets, pricing strategy, segment prioritization | Guides acquisition spend; highlights high-value segments |
Churn Rate (Customer & Revenue) | Low–Medium, straightforward calc, needs segmentation | Customer records, revenue tracking, cohort analysis | Rate of customer/revenue loss; early warning on retention | Retention programs, product-market fit checks, CS focus | Actionable indicator; directly impacts LTV and growth trajectory |
Feature Adoption & Product Engagement | High, requires event tracking and behavioral analysis | Product instrumentation, analytics platform, segmentation | Depth of usage, feature value signals, churn predictors | Product roadmap, onboarding optimization, identifying expansion levers | Predicts retention and expansion; guides product investment |
Free-to-Paid Conversion Rate | Low–Medium, funnel tracking and cohort analysis | Signup/trial events, billing data, traffic attribution | Efficiency of product-led conversion from free to paid | Freemium/trial models, onboarding experiments, growth optimization | Lowers CAC; provides direct feedback on product-market fit |
Customer Health Score & Expansion Potential | High, composite model and ongoing validation | Integrated usage, support, NPS, revenue signals, modeling | Predictive churn risk and expansion opportunities; prioritization | Customer success prioritization, enterprise account management | Enables proactive retention and targeted expansion outreach |
North Star Metrics by Business Model
The eight metrics above are all legitimate north stars — but not all of them are the right starting point for every B2B SaaS business. Which one you prioritise depends on how your product creates and captures value. The biggest mistake founders make is picking a metric because it sounds sophisticated rather than because it maps to their actual GTM motion.
Business model | Primary north star | Supporting metrics | Why |
|---|---|---|---|
Product-led growth (PLG) | Free-to-Paid Conversion Rate | Feature adoption rate, time-to-value, activation rate | Revenue follows product value. Conversion rate tells you whether the product is delivering that value fast enough for users to pay. |
Sales-led growth (SLG) | Net Revenue Retention (NRR) | CAC payback period, MRR growth, expansion revenue | Sales cycles are expensive. NRR tells you whether closed accounts are growing — which determines whether the CAC investment compounds or leaks. |
Usage-based pricing | Feature Adoption + Expansion MRR | Usage growth per account, NRR, power-user depth | Revenue scales with usage. Tracking which features drive the most usage growth tells you where to invest product resources for the highest revenue return. |
Vertical SaaS | Customer Health Score | Churn rate, NPS, support ticket volume, feature breadth per account | Vertical SaaS wins through deep retention, not broad acquisition. Health score catches at-risk accounts before they churn — which matters more in a defined TAM where every customer is hard to replace. |
Infrastructure / API-first | Monthly Active Integrations / API Call Growth | CAC, expansion MRR, NRR | Stickiness comes from being embedded in customer workflows. Measuring active integration depth tells you how embedded you are — and how hard you'd be to rip out. |
Marketplace / network-effect SaaS | Gross Merchandise Value (GMV) or liquidity rate | Supply/demand balance, take rate, repeat transaction rate | Network-effect businesses live and die by liquidity — the percentage of supply that finds demand. Standard SaaS metrics undercount the real health signal. |
North Star Metrics by Company Stage
Your north star should also evolve as your company scales. A metric that's actionable at $50K ARR may be meaningless at $5M ARR — and vice versa. Here's a framework for which metrics deserve the most attention at each stage:
Stage | ARR range | Primary focus | Key north star | What to ignore |
|---|---|---|---|---|
Pre-PMF | $0–$500K | Does the product create enough value that users come back? | Feature adoption rate + Free-to-Paid conversion | NRR (not enough cohort history), CAC (deal volume too low to be statistically meaningful) |
Early growth | $500K–$3M | Is the unit economics story valid at small scale? | CAC payback period + MRR growth rate | LTV models (too early), complex health scores (not enough data) |
Scaling | $3M–$15M | Are we retaining and expanding revenue efficiently? | NRR + churn rate (segmented by cohort and plan) | Raw MRR growth (without NRR context it's misleading), vanity engagement metrics |
Growth | $15M–$50M | What's the ceiling on our best customer segment? | LTV:CAC ratio + expansion MRR by segment | Aggregate churn (use cohort churn instead), simple conversion rate |
Late-stage / pre-IPO | $50M+ | Is growth efficient and durable? | Rule of 40 (growth rate + profit margin), NRR, gross margin | Feature adoption (should be optimised by now), conversion rate (not the bottleneck) |
The trap most founders fall into is tracking too many metrics simultaneously — a dashboard with 15 KPIs is just noise with better formatting. Pick one primary north star per stage, track two or three supporting metrics, and make the rest available for deep dives when something looks off.
If you're building this tracking infrastructure manually in spreadsheets, you're adding lag at every layer. Statspresso's AI Data Chat connects directly to your SaaS data sources — ask "what's our NRR by cohort for accounts signed in Q3?" and get the answer in seconds, not a ticket queue.
From Metrics to Movement Find and Track Your North Star
Choosing a North Star Metric isn't a branding exercise. It's a decision about what kind of company you're building and what evidence you trust most. The right metric gives your team focus. The wrong one gives you alignment theater.
The pattern is usually straightforward. Early-stage SaaS companies should lean toward activation, engagement, or feature adoption. Those metrics tell you whether users are finding value at all. Growth-stage companies can move toward MRR, NRR, churn discipline, and LTV because they have enough customer history to support those views.
That said, no serious operator should worship one metric in isolation. MRR without churn context is misleading. Conversion without activation depth is fragile. Engagement without retention is vanity with better branding.