Tags: commerce analytics concept
Product-Qualified Leads
Date: 2026-09-27
A lead qualified by what someone has done in the product, not by what they filled in on a form. It only exists where people can use the product before paying — and it’s only as good as the behaviour it’s defined on, which has to be checked against who actually buys.
A product-qualified lead (PQL) is a user of a free or trial product whose in-product behaviour suggests they’re ready to buy or to talk to sales.
It sits in the place a marketing-qualified lead (MQL) would, but qualifies on evidence of use rather than evidence of interest — Lead Funnel Stages.
MQL versus PQL
| MQL | PQL | |
|---|---|---|
| Signal | Downloaded content, attended a webinar, visited pricing | Invited teammates, hit a usage limit, connected an integration |
| What it shows | Interest in the topic | Value from the product |
| Requires | A form | A free tier or trial — Free Trial vs Freemium |
| Typical failure | Content consumers who were never buyers | Heavy users who’ll never pay (students, hobbyists) |
The term sits in the product-led growth (PLG) vocabulary — companies where the product itself does most of the acquisition and selling. OpenView’s Blake Bartlett is credited with coining “product-led growth” around 2016. [CHECK: who first used “product-qualified lead” — no clear attribution found.]
Defining one
A PQL is a rule over events. It combines fit (who they are) and usage (what they’ve done):
PQL RULE — example, for a team collaboration tool
fit: company email domain (not gmail/hotmail)
AND company size ≥ 10 (from enrichment)
usage: ≥ 3 teammates invited in the first 14 days
AND ≥ 1 project with activity on 5+ separate days
OR hit the free-plan limit
USERS, DAY 14 PQL?
user domain invites active_days hit_limit
u1 acme.co.uk 4 7 no ✓ fit + usage
u2 gmail.com 6 9 yes ✗ no fit — maybe a self-serve upgrade, not a sales lead
u3 widgets.com 0 1 no ✗ signed up, didn't use
u4 widgets.com 1 3 yes ✓ hit the limit
Notice u2: heavy use, no fit. That’s a self-serve customer or a non-payer, not someone sales should call.
Finding the right behaviour
Start from customers who paid, and look backwards at what they did first — the same analysis that finds an activation event — Activation and Time to Value.
- Compare paying and non-paying users on candidate behaviours in their first week or two
- Prefer behaviours that are costly to fake — inviting colleagues, connecting data — over clicks
- Beware circularity. “Visited the billing page” predicts paying because people who’ve decided to pay go there. It’s a symptom of the decision, not an early signal of it — Leading and Lagging Indicators
- Correlation, not cause. A behaviour that predicts conversion won’t necessarily cause it if you push people into it — Correlation and Causation
Operating it
- Route PQLs by account value. Small ones get an automated upgrade prompt; large accounts get a person
- Speed still matters — Speed to Lead
- Re-check the rule against outcomes each quarter. PQL-to-paid rate falling means the rule has drifted from what buyers do
- Instrument it properly. The rule is only as reliable as the events it reads — Event Taxonomy Design, Identity Stitching for users who sign up on one device and invite from another