TL;DR
- Attribution Persistency: Ensure UTM parameters persist from anonymous browsing to identified signup by configuring the `person_profiles` merge logic.
- The 'Aha Moment' Pipeline: Move beyond simple login tracking. Instrument technical milestones (e.g., `first_api_call_completed`) to calculate precise Time-to-Value (TTV).
- PQL Scoring Engine: Build automated Product Qualified Lead alerts using HogQL to identify "Viral Pockets" within organizations before Sales manually intervenes.
- The Reverse ETL Bridge: Use Census or Hightouch to pipe behavioral product signals back into your CRM, enabling "Contextual Outbound" for your sales team.
- Tool Consolidation: Replacing Mixpanel, Optimizely, and Segment with a unified PostHog stack reduces technical debt and integration latency by 90%.
1. Solving Attribution Persistency
The #1 failure in PLG analytics is losing the "Marketing Origin" of a user. If a user clicks an ad, browses for 3 days, and then signs up, many tools treat them as "Direct" traffic. This breaks your CAC (Customer Acquisition Cost) model.
PostHog solves this via **Person Profiles**. By configuring your SDK to persist `initial_utm_source`, you ensure the first touchpoint is locked to the user ID forever, even after they clear their cookies or change devices, provided you call `posthog.identify()` on signup.
2. Building the PQL Scoring Engine
A Product Qualified Lead (PQL) is not just a "high activity user." It is a user or organization that has achieved a specific **Value Density**. We use HogQL to calculate these scores at the `organization` level, identifying when a self-serve team is ready for an Enterprise upgrade.
The Viral Pocket Formula
We look for three technical signals to trigger a PQL alert:
- Velocity: Usage of core features increased >50% week-over-week.
- Breadth: More than 3 unique users within one domain are performing "Creation" events.
- Integration: The organization has successfully connected a 3rd party API (e.g., Stripe or Salesforce).
"The most valuable signal in B2B PLG is not how often one person logs in, but how quickly the 'Organizational Surface Area' of your product is expanding. If usage jumps from 1 user to 5 in 48 hours, that is a PQL Red Alert."
— Jake McMahon, ProductQuant
3. The Reverse ETL Bridge: Analytics to Action
Data that sits in a dashboard is dead. For PLG to work, product data must live where your Sales and Success teams live (HubSpot, Salesforce, Intercom). PostHog's **Data Pipelines** act as the bridge.
| Signal | Technical Trigger | Sales Action |
|---|---|---|
| The 'Aha' Moment | `first_successful_deployment` | Automated "Congrats" + Best Practices guide. |
| Usage Ceiling | `hits_80_percent_of_tier_limit` | Account Executive reaches out for expansion talk. |
| Feature Gating | `clicks_enterprise_only_feature` | Automated "Upgrade to enable" in-app nudge. |
FAQ
How do we handle multi-tenant B2B tracking?
Use **Group Analytics**. By calling `posthog.group('organization', 'org_id_123')`, you aggregate all behavioral data at the account level. This is essential for calculating NRR (Net Revenue Retention) and account-level health scores.
Can we use PostHog for HIPAA-compliant PLG?
Yes. By deploying PostHog on a **Private Cloud** (VPC), you ensure that sensitive behavioral data stays within your compliance boundary while still benefiting from the full PLG suite (flags, experiments, cohorts).
What is the 'Integration Tax'?
It is the 2-4 week delay and engineering cost required to sync data between fragmented tools (e.g., getting Amplitude data into Intercom). A unified stack like PostHog removes this tax, allowing you to launch a retention experiment in hours, not weeks.
Sources
About the Author
Jake McMahon is a PLG & GTM Growth Consultant who has architected unified experimentation engines for Series A-C SaaS companies. He specializes in the technical instrumentation of growth loops and has led the migration of 100M+ events from legacy stacks to PostHog.