Product-Qualified Leads (PQL)
Purpose: define and instrument the signal that a user (or account) has experienced enough product value to be worth a sales/expansion motion. PQL is the measurement backbone of product-led growth (PLG) — it replaces the marketing-qualified-lead (MQL) "filled a form" proxy with "did meaningful work in the product."
Use this when designing a PLG funnel's conversion layer, defining activation→monetization handoffs, or instrumenting which in-product behavior should trigger sales/expansion.
Contents
- PQL vs MQL vs SQL
- Defining the PQL (signal model)
- PQL vs PQA (account-level)
- Instrumentation
- Thresholds & handoff
PQL vs MQL vs SQL
| Lead type | Qualified by | Owner | Weakness it fixes |
|---|---|---|---|
MQL |
marketing engagement (downloads, form fills) | marketing | engagement ≠ intent |
SQL |
sales acceptance after discovery | sales | slow, manual |
PQL |
demonstrated in-product value | product + growth | predicts conversion far better in PLG motions |
PQL does not replace the funnel — it is the conversion event between activation and revenue
in the PLG funnel (Acquisition → Activation → PQL → Conversion → Expansion).
Defining the PQL — Signal Model
A PQL is a scored combination of three signal classes. Avoid single-event PQLs (one click is noise); avoid 20-factor models (unexplainable). Aim for 3-6 weighted signals.
| Signal class | Examples | What it proxies |
|---|---|---|
Activation depth |
reached the aha-moment, completed core workflow N times | value experienced |
Breadth / habit |
used ≥K features, returned on ≥D days in window | stickiness |
Expansion intent |
hit a plan limit, invited teammates, used a gated feature | willingness to pay |
PQL score = weighted sum, with a hard gate on activation (no activation → never a PQL,
regardless of other signals). Define the aha-moment with activation-design.md; define the
underlying events with event-schema.md.
PQL vs PQA (Product-Qualified Account)
- PQL = an individual user crosses the threshold (self-serve / prosumer motions).
- PQA = an account crosses it — aggregate signals across all users in the org (seat count, cross-team adoption, account-level limit hits). Use PQA for B2B sales-assist motions where the buyer ≠ the active user.
Model PQA as a roll-up of PQL signals plus account-only signals (domain, seats, billing tier).
Instrumentation
- Every PQL signal must map to a tracked event with stable naming (
event-schema.md) — a PQL model built on ad-hoc events silently drifts as the product changes. - Compute the score in the warehouse / product-analytics tool, not in the app, so weights can be re-tuned without a deploy.
- Emit a
pql_reachedevent when the threshold is crossed (timestamped, with the contributing signals) so downstream funnels andexperimentcan measure conversion lift on the PQL cohort.
Thresholds & Handoff
- Calibrate the threshold against historical conversion: pick the score where the conversion-to-paid rate justifies a sales touch (don't guess — backtest on converted users).
- Re-tune quarterly; PQL definitions decay as the product and ICP evolve.
- Handoffs: aha-moment / activation definition →
activation-design.md; event contracts →event-schema.md; PQL-cohort conversion lift →experiment; retention of the PQL cohort →growth; PQL→revenue dashboards →dashboard-spec.md/revenue-analytics.md.