How to Measure Product-Led Growth: The Metrics That Actually Matter
PLG sounds simple until you're measuring the wrong things. This guide covers the specific behavioral metrics — activation milestones, NRR, feature adoption depth, viral coefficient — that separate PLG teams that compound from those that stall.
Share
How to Measure Product-Led Growth: The Metrics That Actually Matter
Product-led growth sounds simple: let the product do the selling. But most teams implementing PLG end up measuring the wrong things, optimizing for vanity metrics, and wondering why their self-serve funnel isn't compounding the way it should.
The problem isn't PLG itself — it's that PLG requires a fundamentally different measurement stack than sales-led growth. When a sales rep closes a deal, you know exactly when, why, and who. When a product closes a deal, you have to reconstruct that from behavioral signals — and most teams haven't instrumented for that.
This guide covers the specific metrics that separate PLG teams that compound from PLG teams that stall.
Why Standard SaaS Metrics Miss the Point in PLG
In a sales-led model, the funnel is linear: lead → opportunity → demo → close. The metrics match: MQLs, SQLs, win rate, sales cycle. You measure hand-offs between humans.
In PLG, there are no hand-offs. The funnel is behavioral: awareness → sign-up → activation → habit → expansion. Every transition is driven by what the user does — or doesn't do — inside the product. You're measuring product behaviors that predict revenue, not pipeline stages that contain it.
This is why teams that copy their sales-led metrics into a PLG context end up confused. "We have 10,000 sign-ups" means nothing if 9,800 of them never experienced value. "Our trial-to-paid conversion is 8%" means nothing if you don't know what the 8% did differently inside the product.
PLG metrics are behavioral, predictive, and compounding. They answer a different question: not "how many people are in which stage?" but "what behaviors predict retention, expansion, and referral?"
The Four Layers of PLG Measurement
Layer 1: Acquisition Quality
In PLG, acquisition quality matters more than acquisition volume. A thousand sign-ups from a billboard campaign are worth less than ten sign-ups from a blog post targeting your exact ICP — because the latter actually use the product.
S
Specky Team
Writing about AI-native product development at Specky.
Sign-up to activation rate by source: Break down your activation rate by acquisition channel. You'll almost certainly find that organic search, product comparison pages, and word-of-mouth convert at 2-3x the rate of broad paid campaigns. This is your signal to reallocate marketing spend toward content and community, not more top-of-funnel ads.
Time to first value by ICP segment: How long does it take a user from your ideal customer profile to reach their first "aha" moment? The tighter this window, the stronger your PLG motion. If ICP users take 3 days to activate and non-ICP users take 10 (and most churn before they get there), you've just identified both a targeting problem and a product friction problem.
Qualified product-led signups (PQLs): Not all sign-ups are created equal. A PQL is a sign-up that matches your ICP criteria (company size, role, industry) AND has completed at least one activation step. This is the metric your growth team should be optimizing — not raw sign-ups.
Layer 2: Activation
Activation is the most important — and most misunderstood — metric in PLG. Most teams define activation wrong.
Activation is not "completed onboarding." It's not "logged in twice." It's not "watched the intro video."
Activation is the point at which a user has experienced enough value that they are likely to become a long-term retained user. It's predictive, not descriptive.
To find your real activation milestone, run a cohort analysis: take users who are retained at 30 days, and find the earliest product behavior that the retained cohort completed at a significantly higher rate than churned users. That behavior — not the one your product team assumed — is your real activation milestone.
Metrics:
Activation rate (to the real milestone): The percentage of sign-ups who complete the behavior that actually predicts retention. If you haven't run the cohort analysis to find this milestone, you're measuring the wrong thing.
Time-to-activation: How many minutes, hours, or days from sign-up to activation milestone? In PLG, speed matters. Every hour of friction between sign-up and first value is churn risk. The best PLG products get users to value in a single session.
Activation funnel drop-off by step: Where specifically do users fall out before reaching activation? This is your product roadmap for onboarding improvements. A 60% drop-off at "connect your first integration" is a different problem than a 60% drop-off at "invite a teammate."
Layer 3: Engagement and Habit
Retention is the engine of PLG compounding. An activated user who churns in month 2 contributes nothing to word-of-mouth, expansion, or community. Engagement metrics tell you whether you're building habits or just novelty.
Metrics:
Daily/Weekly active users (DAU/WAU) and DAU/MAU ratio: The DAU/MAU ratio (sometimes called "stickiness") tells you what fraction of your monthly active users are coming back daily. For a daily-use tool like Slack or Notion, 50%+ is excellent. For a weekly tool like a PM workspace, 30%+ WAU/MAU is strong. The number matters less than the trend — and the trend matters less than whether your power users are daily and your at-risk users are monthly.
Feature adoption depth: What percentage of activated users use N or more core features? A user who uses only one feature is at much higher churn risk than one who uses five. Deep feature adoption is a proxy for switching cost — the more of your product they rely on, the harder it is to leave.
Breadth of use within accounts: In B2B PLG, team-level adoption matters more than individual adoption. A single power user at a company doesn't translate to expansion or long-term retention. Monitor the percentage of accounts where 3+ seats are active — that's the signal that PLG has gone viral inside an organization.
Session frequency and depth: Not just how often users log in, but what they do when they're there. A user who logs in daily and completes 3+ meaningful actions is very different from one who logs in to check a notification and leaves. Build event-based engagement scores that reflect actual value consumption, not just presence.
Layer 4: Expansion and Virality
This is where PLG compounding actually comes from — and where most teams leave money on the table.
Metrics:
Net revenue retention (NRR): The single most important PLG health metric. NRR measures what percentage of revenue you retain from existing customers after accounting for churn, downgrades, upgrades, and expansion. An NRR above 100% means your existing customer base is growing even without new logo acquisition — the hallmark of true PLG compounding. Benchmark: 110%+ is good; 120%+ is exceptional; below 100% means you're on a treadmill.
Product-qualified leads for expansion (PQL expansions): Users in existing accounts who are hitting usage limits, inviting teammates at high frequency, or using advanced features at high rates are expansion signals. These are warm leads for your account management or self-serve upsell flow — and they're usually much easier to convert than new logos.
Viral coefficient (K-factor): How many new sign-ups does each existing user generate through invites, shares, or referrals? A K-factor above 1 means your product is growing on its own. Most B2B SaaS products have K-factors between 0.1 and 0.3 — positive virality, but not explosive. Build sharing mechanics into your activation flow (not as an afterthought) to move this number.
Time-to-expansion: How long after sign-up does a user or account typically expand? If the median expansion happens at month 3 but you're doing outreach at month 6, you're missing the window. PLG expansion should be triggered by product behavior, not by a calendar invite.
The PLG Metric Stack in Practice
Putting it together, here's what a PLG dashboard should show at any given moment:
Weekly pulse (tactical):
New PQL sign-ups this week vs. last week
Activation rate (7-day cohort)
Time-to-activation (median, this cohort vs. 30-day average)
DAU/WAU trend
Monthly review (strategic):
Activation rate by acquisition source (find the channels that produce real users)
30/60/90-day retention by activation milestone cohort
Feature adoption depth distribution
Breadth of use within accounts (% of accounts with 3+ active seats)
NRR
PQL expansion pipeline
Quarterly deep-dive:
Re-run the cohort analysis: has the real activation milestone changed?
Segment all metrics by ICP vs. non-ICP to verify you're serving the right customers
Viral coefficient trend
Time-to-expansion by acquisition cohort
The Most Common PLG Measurement Mistakes
Mistake 1: Using the wrong activation milestone
If your "activation" metric is "completed onboarding checklist," you're measuring form completion, not value. Do the cohort analysis. Find the behavior that actually predicts retention. Rebuild around that.
Mistake 2: Measuring monthly active users without segmentation
MAU is a vanity metric unless you know which MAUs matter. A user who logs in once a month to check a notification is not "active" in any meaningful sense. Segment by engagement depth, ICP fit, and account expansion potential.
Mistake 3: Ignoring team-level signals in B2B
Individual user metrics miss the point in B2B. A company where one person uses your product intensely is at churn risk the moment that person leaves. Monitor account-level activation, adoption breadth, and seat expansion as first-class metrics.
Mistake 4: Treating NRR as a finance metric
NRR is a product metric. If your NRR is declining, it means your product is not delivering enough ongoing value to justify renewal and expansion. That's a product problem, not an account management problem. Own it accordingly.
Mistake 5: Building the viral loop as an afterthought
If your invite/share mechanic is buried in Settings and only triggered by a monthly email reminder, your K-factor will be near zero. PLG virality is engineered — it's in the activation flow, the first share moment, the collaborative feature. Plan it before you build, not after.
How Specky Instruments PLG Measurement
Specky's feature adoption view is built specifically for PLG teams who need to close the loop between product behavior and strategic decisions. Instead of tracking events in isolation, Specky links adoption signals to the graph nodes they affect — so when feature usage spikes among a segment, it surfaces automatically alongside the interviews, tickets, and insights that explain why.
The feature adoption view shows:
Adoption rate by feature, by cohort, and by ICP segment
Usage depth curves (how intensely do adopters use the feature?)
At-risk accounts (high ICP fit, low recent engagement — the expansion opportunity hiding in plain sight)
Correlation between feature adoption and NRR (which features do retained customers use that churned ones didn't?)
Most PLG teams have the data. The problem is that it lives in PostHog, Mixpanel, and a spreadsheet — disconnected from the product decisions it should be informing. Connecting them is what makes measurement compound into strategy.
Quick Reference: PLG Metric Stack
Layer
Metric
What It Tells You
Acquisition
Sign-up to activation by source
Which channels produce real users
Acquisition
Time to first value by ICP
Where ICP friction lives
Activation
Activation rate (real milestone)
Whether users reach the value threshold
Activation
Time-to-activation
Speed of the first value delivery
Activation
Funnel drop-off by step
Where to focus onboarding investment
Engagement
DAU/WAU ratio
Whether you're building daily habits
Engagement
Feature adoption depth
Proxy for switching cost
Engagement
Breadth of use per account
Team virality signal
Expansion
NRR
Overall PLG health
Expansion
PQL expansion pipeline
Self-serve upsell opportunity
Expansion
Viral coefficient (K-factor)
Organic growth multiplier
Measure these. Link them to product decisions. That's how PLG compounds.