The Analytics Setup for Product-Led Growth Companies (From Signup to Expansion)

Product-led growth changes what analytics needs to do. In a sales-led model, the handoff is clear: marketing generates leads, sales qualifies them, deals close, and analytics reports on pipeline stages. In a product-led model, the product is the sales team. Users sign up, experience value, invite teammates, hit usage limits, and upgrade themselves. The "pipeline" is not in your CRM. It is in your product. And if your analytics stack was built for a sales-led motion, you are measuring the wrong things entirely.

PLG analytics needs to answer questions that traditional analytics does not even ask. Which free-tier behaviors predict conversion to paid? At what usage threshold do teams naturally expand? Which onboarding paths lead to activation versus abandonment? How long does the average self-serve upgrade take, and what accelerates it? These are not marketing questions or sales questions. They are product questions that have direct revenue implications. And answering them requires an analytics setup that most PLG companies do not have, even well-funded ones.

Why Traditional Analytics Fails for PLG

Traditional B2B analytics was built around the marketing funnel: visitor, lead, MQL, SQL, opportunity, closed-won. Each stage has clear criteria, and the system measures how efficiently people move through these stages. This model assumes that marketing generates interest, sales converts interest into revenue, and the product is what gets delivered after the deal closes. In PLG, this model is backwards.

In PLG, the product generates interest, demonstrates value, and converts users to paying customers. Sales (if it exists at all) enters the picture for expansion, enterprise deals, or accounts that need help. The funnel is not marketing-to-sales. It is product-to-revenue. And the metrics that matter are fundamentally different.

The MQL Trap

Many PLG companies still use MQLs as their primary lead signal because that is what their CRM and marketing automation tools were designed to produce. A user downloads a whitepaper, attends a webinar, or visits the pricing page three times, and the system flags them as an MQL for sales follow-up. The problem is that in PLG, content consumption is a weak signal. A user who reads your blog post but never signs up is less valuable than a user who signed up for a free plan, created three projects, and invited two teammates. The second user has demonstrated real product engagement, but the MQL system never sees this because product usage data does not flow into the marketing automation platform.

The result is a sales team chasing content downloaders while ignoring highly engaged free users who are one nudge away from upgrading. PLG analytics fixes this by replacing the MQL with the product-qualified lead (PQL), a signal based on in-product behavior rather than marketing engagement.

The Five Stages of PLG Analytics

The PLG analytics framework tracks five stages: acquisition, activation, engagement, monetization, and expansion. Each stage has specific events to track, metrics to measure, and questions to answer. The stages are sequential in theory but overlapping in practice, because users move through them at different speeds and sometimes skip stages entirely.

Stage 1: Acquisition Analytics for PLG

PLG acquisition is different from traditional B2B acquisition because the goal is not a lead form submission. It is a product signup. This changes what you measure and how you optimize.

Beyond Signup Volume

Total signups is a vanity metric in PLG. What matters is signups that activate. If you get 10,000 signups per month but only 15% reach activation, your effective acquisition is 1,500, not 10,000. Break signups down by source and measure the activation rate for each source. You will typically find that some channels produce high signup volume with low activation (paid social, content syndication) while other channels produce lower volume with much higher activation (organic search for product-related queries, referrals from existing users, community-driven signups).

The activation-adjusted acquisition cost is the PLG replacement for traditional CAC. If a channel costs $50 per signup but only 10% activate, the real cost per activated user is $500. A different channel might cost $80 per signup with a 40% activation rate, producing activated users at $200 each. The second channel is 2.5x more efficient despite the higher per-signup cost. You cannot see this without analytics that connect acquisition source to in-product activation behavior.

Signup Flow Optimization

Track every step of the signup flow as a funnel: landing page view, signup page view, form start, email submission, email verification, onboarding start. Measure the drop-off at each step and segment by device, source, and user type. Common PLG findings include: mobile signup completion rates are 40-60% lower than desktop (fix: simplify the mobile form), social login (Google, GitHub) has 2-3x higher completion than email signup (fix: make social login the primary option), and email verification is the single largest drop-off point (fix: use magic links or delay verification until after initial product experience).

Stage 2: Activation Analytics

Activation is the most important stage in PLG analytics. It is the bridge between signup and retention. Users who activate retain. Users who do not activate churn. Everything else is downstream of activation.

Defining Your Activation Metric

Your activation metric is the specific in-product behavior that correlates most strongly with long-term retention. For Slack, it is a team sending 2,000 messages. For Dropbox, it was saving a file in a Dropbox folder. For a project management tool, it might be creating a project and adding at least one task. The activation metric is discovered through data analysis, not assumed. Analyze the behaviors of users who retain after 90 days and compare them to users who churned within 30 days. The behaviors that are most different between these two groups are your activation candidates.

Test multiple candidate metrics. "Created a project" might have a 60% correlation with 90-day retention. "Created a project AND invited a teammate" might have an 85% correlation. "Created a project, invited a teammate, AND used a template" might have a 92% correlation. The activation metric with the highest retention correlation and a reasonable percentage of users achieving it is your winner. If only 2% of users achieve the metric, it is too restrictive. If 90% achieve it, it is not predictive enough. Aim for a metric that 30-50% of signups achieve and that has a 70%+ correlation with retention.

Time-to-Value

Time-to-value (TTV) measures how long it takes a new user to reach the activation milestone from the moment they sign up. Shorter TTV correlates with higher activation rates. If your average TTV is 5 days, users need to stay motivated for 5 days before they experience enough value to stay. Every day of delay is an opportunity for the user to get distracted, forget about your product, or try a competitor. The best PLG products achieve TTV in minutes: Canva lets you create a design immediately, Loom lets you record a video immediately, Notion gives you a workspace immediately. Track TTV as a distribution, not an average. The average might be 3 days, but if 40% of users activate in under an hour and 30% never activate at all, the average tells you nothing useful.

Stage 3: Engagement Analytics

After activation, engagement analytics tracks whether users continue finding value in the product. Engagement is the leading indicator of both retention and expansion. Users who are deeply engaged retain and eventually upgrade. Users whose engagement is declining are on a path to churn.

The DAU/MAU Ratio

The DAU/MAU ratio (daily active users divided by monthly active users) is the standard measure of product stickiness. A ratio of 50% or higher means the average user comes back every other day, which is exceptional for B2B SaaS. Most B2B products land between 15-30%. The ratio matters more than absolute DAU or MAU because it measures habit formation. A product with 10,000 MAU and a 40% DAU/MAU ratio has stronger retention dynamics than a product with 50,000 MAU and a 10% ratio, even though the second product has more users.

Track DAU/MAU over time and by cohort. If the ratio is declining for recent cohorts, the product is becoming less sticky for new users, possibly because core features are being diluted, onboarding is degrading, or the competitive landscape is pulling users away. If the ratio is increasing, your recent product improvements are driving more habitual usage.

Feature Adoption Breadth and Depth

Track how many distinct features each user engages with (breadth) and how intensively they use each feature (depth). Users who use 5+ features retain at significantly higher rates than users who rely on a single feature. Single-feature users are vulnerable to any competitor that does that one thing better. Multi-feature users have built workflows around your product and face high switching costs.

Build a feature adoption matrix that shows, for each feature, the percentage of users who have discovered it, tried it once, used it repeatedly, and made it part of their regular workflow. This matrix reveals features that are well-discovered but abandoned after first use (value problem or UX friction), features that are heavily used by a small group but unknown to the majority (discoverability problem), and features that see broad, deep usage (your product's core value).

Stage 4: Monetization Analytics

Monetization in PLG is primarily self-serve. Users hit a limit, see a gate, or decide they want premium features, and they upgrade themselves. Analytics needs to track this entire self-serve conversion journey, from the first paywall encounter to the completed payment.

Paywall and Limit Encounters

Track every time a free user encounters a paid feature gate or usage limit. This event is the start of the monetization funnel. What matters is not just that they hit a limit, but how they respond: do they dismiss the upgrade prompt, do they view the pricing page, or do they start the upgrade flow? Segment limit encounters by the specific limit hit (user cap, storage cap, feature gate, API limit) and measure the conversion rate for each. You will find that some limits drive upgrades at 10x the rate of others. Focus your monetization design on the limits that naturally convert rather than adding more gates that frustrate users.

Self-Serve Conversion Funnel

The self-serve upgrade funnel is: limit encounter or feature discovery, pricing page view, plan selection, payment information entry, payment completion. Track each step and measure drop-off. Common findings: users who compare multiple plans on the pricing page convert at higher rates than users who view only one plan (they are evaluating seriously, not browsing), annual billing selection is a strong buying signal (users who select annual billing complete payment at 2x the rate of monthly selectors), and payment form abandonment is often caused by unexpected charges or unclear billing terms rather than price itself.

Stage 5: Expansion Analytics

Expansion revenue (upgrades, seat additions, usage increases) is the growth engine of PLG companies. Net dollar retention above 120% means the company grows even without acquiring new customers. Analytics for expansion focuses on identifying natural expansion signals and measuring the effectiveness of expansion prompts.

Natural Expansion Triggers

Track the events that naturally precede expansion. When an account adds its 5th team member on a plan with a 5-user limit, that is a natural expansion trigger. When a team uses 80% of its storage allocation, that is a natural expansion trigger. When an account starts using a feature that is available on a higher plan, that is a natural expansion trigger. Map every limit and gate in your product, track how often accounts approach each one, and measure the expansion conversion rate for each trigger.

The most sophisticated PLG analytics setups build expansion propensity models using these triggers. An account that has hit 3 usage limits, has 10+ active users, and has been on the current plan for 6+ months has a high expansion propensity. This signal can route to the sales team for proactive outreach or trigger an automated in-product upgrade prompt. The key is using behavioral data rather than firmographic data to predict expansion. Company size and industry are weak predictors. Product usage patterns are strong predictors.

Revenue Attribution in PLG

PLG revenue attribution is more complex than traditional attribution because the "conversion" happens inside the product, often weeks or months after the initial acquisition touch. A user might discover your product through a blog post, sign up two weeks later via a direct visit, activate over the following month, and upgrade three months after that. The blog post, the direct visit, the activation experience, and the upgrade prompt all contributed to the revenue. PLG attribution models need to weight both marketing touches and in-product experiences.

Building Product-Qualified Leads (PQLs)

The PQL is the PLG equivalent of the MQL, but instead of being defined by content engagement, it is defined by product behavior. A PQL is a free user or account that has demonstrated enough product engagement to warrant sales outreach. The PQL model converts behavioral data into a sales signal.

Defining PQL Criteria

Build PQL criteria from two dimensions: product engagement and firmographic fit. Product engagement includes activation status, feature usage breadth, team size, usage frequency, and limit encounters. Firmographic fit includes company size, industry, and role (from signup data or enrichment). The PQL threshold should produce a manageable volume for your sales team and a high enough conversion rate to justify outreach. Start with a strict threshold (high engagement + good fit) and loosen it as you learn.

Score PQLs on a continuous scale rather than a binary yes/no. A user who has activated, invited 3 teammates, hit 2 usage limits, and works at a 500-person company scores higher than a user who has activated but uses the product alone at a 10-person company. The continuous score helps sales prioritize accounts and determines the appropriate outreach approach: high-score PQLs get immediate personal outreach, mid-score PQLs get automated nurture sequences, low-score PQLs stay in self-serve.

The PLG Analytics Stack

The analytics stack for PLG companies needs to span marketing, product, and revenue data. No single tool covers all three. Here is how to build a stack that provides complete visibility across the entire PLG journey.

Product Analytics (Core)

Your product analytics platform (Mixpanel, Amplitude, PostHog, Kissmetrics) is the centerpiece. It tracks events inside your product, enables funnel analysis, cohort analysis, and behavioral segmentation. Choose a platform that supports both marketing-site events and in-product events so you can build funnels that span the entire journey from first website visit to product activation. Avoid platforms that require you to stitch together separate marketing analytics and product analytics tools.

Customer Data Platform (Routing)

A CDP (Segment, RudderStack) collects events from all sources and routes them to all destinations. In PLG, the CDP connects your marketing site, product application, mobile app, and backend systems into a unified event stream. Without a CDP, you end up with siloed data in each tool and no way to build a complete user journey. The CDP also enables identity resolution across touchpoints: connecting the anonymous website visitor to the signed-up user to the paying customer.

Data Warehouse (Truth)

The data warehouse (BigQuery, Snowflake) is where all data lives long-term. Product events, CRM records, billing data, and marketing data all flow into the warehouse. This is where you build the PQL model, calculate lifetime value by cohort, and analyze expansion patterns. The warehouse enables analyses that no individual tool supports because it combines data across tools.

Reverse ETL (Activation)

Reverse ETL (Census, Hightouch) pushes warehouse insights back into operational tools. PQL scores flow from the warehouse to Salesforce. Churn risk signals flow to your customer success platform. Usage-based billing data flows to your billing system. Without reverse ETL, your warehouse insights are trapped in dashboards. With it, they drive automated workflows, sales outreach, and product experiences.

Implementation Roadmap: 12 Weeks to PLG Analytics

PLG Metrics That Matter

The standard SaaS metrics (ARR, CAC, LTV, churn) still matter in PLG, but they need to be supplemented with PLG-specific metrics that measure the health of the product-led motion. Here are the metrics that should be on your PLG dashboard.

Product-led growth is not just a distribution strategy. It is a fundamentally different way of building and growing a software company, and it demands a fundamentally different analytics setup. The companies winning at PLG are not the ones with the best product or the most funding. They are the ones that understand their users deeply enough to remove friction, accelerate value delivery, and let the product do the selling. That understanding comes from analytics that tracks the entire journey, from the first anonymous visit to the expansion revenue event, as one continuous story. Build that story, and the growth follows.