Product-Led Growth Metrics: Which Ones Actually Predict Expansion

What Makes PLG Metrics Different From Traditional SaaS Metrics

Product-led growth shifts the primary sales motion from a human to the product itself. In a sales-led motion, the information asymmetry between buyer and seller is resolved by a sales rep. In a PLG motion, that asymmetry is resolved by the user's direct experience with the product. The metric system must reflect this shift — it needs to instrument the product experience, not just the commercial outcome.

Traditional SaaS growth metrics — monthly recurring revenue (MRR), churn rate, pipeline velocity — describe what the revenue machine produced. They are output metrics. PLG requires a parallel layer of input metrics that describe what the product experience produced, because the product experience is now the primary revenue driver.

This is where most teams get confused. They add product analytics tooling to an existing SaaS metric stack and assume the combination gives them PLG visibility. It does not. Appending product data to a sales-led metric framework does not produce a PLG measurement system. It produces a more complicated version of the same output-only view.

The meaningful distinction is timing. Output metrics describe a completed transaction. Leading product metrics describe an in-progress experience that will become a transaction — or not. Getting the timing right is the foundational principle of PLG measurement.

Tier 1: Leading Indicators That Compound Into Expansion

Leading indicators are metrics that shift reliably before a revenue event. They do not cause expansion directly — a user reaching a usage threshold does not automatically generate an invoice. But they are the most reliable upstream proxies available for predicting whether expansion will occur.

Activation Depth

Activation depth measures how far into the product's core value loop a new user travels within a defined early window — typically the first 7 or 14 days. It is not a binary "did they activate" flag. It is a scored spectrum that maps which specific actions correlate most strongly with downstream retention and expansion.

The OpenView Partners 2024 SaaS Benchmarks report found that companies with clearly defined activation milestones retain users at measurably higher rates than those relying on login-based proxies (OpenView Partners SaaS Benchmarks, 2024). The reason is structural: a defined activation milestone forces the team to specify what value delivery actually looks like, which forces the product itself to get better at delivering it.

Activation depth is a leading indicator because it changes within the first two weeks of a user's lifecycle. Expansion and churn events happen months later. The gap is the intervention window.

The insight: Activation depth is not a metric you report — it is a metric you operationalize. If it does not change what the product team builds next sprint, it is not functioning as a leading indicator.

Feature Adoption Velocity

Feature adoption velocity measures the rate at which users discover and repeatedly use the features that generate the most retention value — sometimes called "sticky features" or "core habit loops." Velocity adds a time dimension that simple feature adoption counts miss.

A user who reaches a sticky feature on day 3 behaves very differently from a user who reaches the same feature on day 21. Both show the same binary adoption. Only the velocity metric captures the difference — and the velocity difference predicts significantly different 90-day retention outcomes.

Feature adoption velocity requires knowing which features are actually "sticky" for your specific user base. This is not something a generic analytics template can answer. It requires running cohort retention analysis segmented by feature usage to identify which features, adopted at which speed, produce the retention curves you want to replicate.

The insight: Feature adoption velocity is the early signal that tells you whether a user is on track to become a retained, expanding account — before any commercial signal is available.

Time-to-First-Value (TTFV)

Time-to-first-value measures the elapsed time between a user signing up and the moment they receive a clear, unambiguous outcome from the product. The definition of "value" is product-specific and must be defined precisely — not as "they completed onboarding" but as "they received output X or achieved outcome Y."

TTFV is a leading indicator because it is the upstream determinant of activation depth. Users with longer TTFV have dramatically higher abandonment rates before ever reaching a sticky feature. Every unnecessary step in the path to first value is a drop-off point — and every drop-off before first value is lost activation that can never recover.

"The single biggest lever in a PLG funnel is almost always time-to-value. Most teams think their activation problem is a user education problem. It is almost always a product friction problem — too many steps between signup and the first 'aha' moment."

— Wes Bush, Product-Led Growth: How to Build a Product That Sells Itself, Product Led Institute (productled.com/book)

Reducing TTFV typically requires cross-functional coordination between product, engineering, and onboarding design. It is not a content problem. It is an architectural problem — how many steps does the product require before it delivers something the user can point to and say "this helped me."

The insight: TTFV is the single most compressible leading indicator in most PLG funnels — and compression directly translates to downstream activation, retention, and expansion rates.

Tier 2: Lagging Indicators That Confirm What Already Happened

Lagging indicators are not useless — they are the definitive scoreboard. NRR tells you whether your PLG motion is compounding. Seat expansion rate tells you whether the product is spreading within accounts. PQL conversion rate tells you whether your product-to-sales handoff is working. But lagging indicators confirm outcomes; they do not predict them. By the time they move, the leading indicators moved months ago.

Net Revenue Retention (NRR)

NRR measures the percentage of revenue retained from an existing customer cohort over a period, including expansion revenue from upgrades and seat additions, minus contraction and churn. An NRR above 100% means existing customers are generating more revenue over time without adding new logos. An NRR above 110% means the product itself is compounding.

For B2B SaaS in the $1–50M ARR range, the KeyBanc Capital Markets Private SaaS Survey has consistently shown median NRR in the 106–112% range for top-quartile performers. Elite PLG companies often sustain NRR above 120%, meaning the installed base is growing faster than churn without proportional sales investment.

NRR is a lagging indicator because it is measured monthly or quarterly, always in arrears. The product behaviors that drive it — feature adoption, activation depth, collaboration spread — happened weeks or months before the NRR figure changes. Watching NRR closely is essential, but trying to manage NRR directly is like trying to steer by looking at the wake behind the boat.

The insight: NRR is your most important output metric and your least useful real-time signal. Manage the Tier 1 inputs; read NRR as the downstream confirmation.

Seat Expansion Rate

Seat expansion rate measures how quickly a product spreads within an account after initial adoption. In bottom-up PLG motions, a single champion adopts the product, generates visible value, and other team members or departments request access. The rate at which this happens within a defined window — say, seats added within 90 days of initial account creation — is the seat expansion rate.

Seat expansion is a lagging indicator because it depends on the initial user first having a strong activation experience, then using the product visibly enough that colleagues notice, then those colleagues requesting access or being invited. Each step takes time, and none of the steps are visible in the seat expansion metric itself.

The leading indicators that precede seat expansion are collaboration-feature adoption and output-sharing behaviors. If your product has a sharing, exporting, or commenting function, and you can measure how quickly new users adopt it, that adoption rate predicts seat expansion before any new seats are created.

The insight: Seat expansion rate is a reliable lagging confirmation of product virality — but you need to instrument the sharing and collaboration behaviors that precede it to have any predictive power.

Product Qualified Lead (PQL) Conversion Rate

A product qualified lead (PQL) is a free or trial user who has reached a defined usage threshold that correlates with a high probability of converting to a paid account. The PQL conversion rate measures what percentage of users who reach that threshold actually convert within a defined window.

PQL conversion is a lagging indicator because PQL status itself is a lagging designation — you can only call a user a PQL after they have already demonstrated the qualifying behavior. The leading indicator is the trajectory of behavior before the PQL threshold: how quickly is this user accumulating usage actions that point toward PQL status?

The insight: PQL conversion rate is most useful as a ratio to track over time, not as an absolute number to benchmark against other companies — your PQL definition is product-specific, and cross-company comparison is structurally misleading.

The Full PLG Metrics Comparison: Signal vs. Noise

The table below applies the signal-before-revenue test to the most commonly tracked PLG metrics. Leading indicators change before revenue events. Lagging indicators change at the same time as or after revenue events. Vanity metrics change without reliable connection to revenue events at all.

Tier 3: Vanity Metrics That Drain Attention Without Predicting Revenue

Vanity metrics are not harmful because they are wrong. They are harmful because they are ambiguous — they can mean multiple contradictory things simultaneously, and they change for reasons that have no connection to revenue outcomes. A dashboard full of vanity metrics creates the experience of visibility without the substance of it.

DAU/MAU Ratio Without Context

DAU/MAU ratio measures how often users return to the product within a calendar month. A higher ratio means users are returning more frequently. This sounds good. But the ratio is structurally ambiguous in ways that matter for PLG.

Consider two users with identical DAU/MAU ratios. User A returns daily because the product automates a tedious task in under two minutes, and they run it every morning. User B returns daily because the product is slow to load, requires manual re-entry of data each session, and takes them multiple visits to complete what should be a single workflow. Both have the same ratio. One is experiencing high value. One is experiencing high friction. The DAU/MAU ratio does not distinguish them.

The consumer app world imports DAU/MAU from social media and gaming, where frequency genuinely correlates with engagement and monetization. In B2B SaaS, the correlation breaks down. A sales analytics tool used twice a week by every account executive in an enterprise account is more valuable — and more expansion-likely — than a tool used daily by one person who has not shared it with anyone.

The insight: DAU/MAU ratio is useful for consumer apps with ad-based monetization. For B2B SaaS with expansion revenue models, it measures the wrong thing at the wrong unit of analysis.

Login Frequency

Login frequency has the same structural problem as DAU/MAU but at the individual user level. It measures authentication events, not value delivery events. In a product with friction — slow loading, session timeouts, required re-authentication — login frequency can actually be inversely correlated with product quality.

The version of this metric that belongs in a PLG dashboard is not login frequency but "value delivery frequency" — how often does a user complete an action that the product's own retention analysis identifies as correlated with long-term retention. That is a different event, instrumented differently, and almost never the same as a login.

The insight: Replace login frequency with action frequency on your defined "sticky" behaviors. The substitution takes engineering effort but produces a metric with actual predictive value.

Average Session Duration

Average session duration is directionally ambiguous in both directions. Long sessions can mean users are deeply engaged with a rich product. They can also mean users are confused, lost in navigation, or struggling to find what they need. Short sessions can mean the product delivers value so efficiently that users complete their task and leave — which is exactly what a well-designed tool should do.

The direction of "good" depends entirely on the product's design intent, and that intent is rarely encoded in how the metric is reported. Without a benchmark specific to the product's task completion model, average session duration generates debate rather than decisions.

The insight: Average session duration is most useful when segmented by user intent — task completion sessions versus exploratory sessions. Without that segmentation, it is noise.

The Signal-Before-Revenue Test: A Practical Filter

The signal-before-revenue test is the fastest way to audit any metric you are currently tracking. The question is simple: does this metric change before a revenue event, at the same time as a revenue event, or without any reliable relationship to a revenue event?

To run the test, pull a cohort of users who expanded (upgraded, added seats, renewed at higher MRR) and a cohort of users who churned. Look at each metric 30, 60, and 90 days before the revenue event. If the metric was meaningfully different between the two cohorts at 60 days before the event, it passes the test. If it only diverges at the event itself — or does not diverge at all — it fails.

This analysis requires product analytics instrumentation and at minimum six months of user data to run cleanly. But the investment is asymmetric: a single cohort analysis can permanently restructure which metrics your team operates on, and that structural change compounds every quarter thereafter.

Building a PLG Metrics Dashboard That Survives a Board Review

A PLG metrics dashboard should be organized into three views with explicit separation of purpose. The primary view — the one that drives sprint planning and product decisions — should contain only leading indicators. The secondary view — the one that drives quarterly strategy and investor reporting — should contain lagging indicators. Vanity metrics should not appear in either view.

Operationally, the primary view should answer: which user cohorts are underperforming on activation, where in the feature adoption sequence are users dropping off, and how is TTFV trending across the last 90 days? If those questions cannot be answered from the primary view, the dashboard is not doing its job.

The secondary view should answer: is NRR improving, what is driving seat expansion, and are PQL conversion rates trending in the right direction? These are confirmation questions — they tell you whether the work on the leading indicators is paying off.

A board or investor audience will always ask about NRR and churn. Your job is to explain what leading indicators you are managing to produce those outcomes next quarter, not this one.

Connecting Activation, Monetization, and Expansion in One System

The most common failure in PLG measurement is treating activation, monetization, and expansion as separate metric systems that happen to share a product. They are not separate. They are sequential phases of one compounding system, and the metrics for each phase must be designed to connect to the next.

Activation depth in phase one should be defined using the same events that predict conversion in phase two. Conversion in phase two should track the same feature-usage patterns that predict seat expansion in phase three. If the metric definitions across phases are inconsistent, the system cannot compound — it can only report each phase separately, which is what most teams are doing when they feel like they have "a lot of data but not enough insight."

ProductQuant was built specifically to close this connection gap for B2B SaaS teams at $1–50M ARR. The work starts with a diagnostic that maps which activation events in your current product data predict downstream expansion — and builds the measurement architecture from that empirical foundation, not from generic PLG templates.