What exactly is a PQL, in plain terms?

**A PQL is a product-qualified lead: a free user whose behaviour inside your product shows they are ready to buy.** A normal marketing lead is someone who filled a form or downloaded a guide. A PQL is different: they have actually used your product and hit the moments that predict payment. **Simple test:** if a free user has done the handful of actions that your paying customers all did before upgrading, they are a PQL. That is the whole idea.

Do I need machine learning to build this?

**No. Start with a simple points model.** Assign points to each meaningful action, add them up, and flag anyone above a threshold. This gets you 80% of the value with almost none of the complexity. Machine learning only earns its place once you have thousands of scored users with known outcomes and a stable points model to improve on. Do not start there.

How many events should I track to begin?

**Three to five, no more.** The temptation is to track everything. Resist it. A firehose of 200 events is data you will never analyse and a maintenance burden. **-** One activation event (they completed setup) **-** One or two aha events (they felt the value) **-** One expansion event (invited a teammate or hit a limit) You can always add events later once the basic score is working.

Is tracking user behaviour a privacy risk?

**It can be, so handle it properly.** Product analytics is normal and legitimate, but you have real obligations: **-** Never put passwords, payment details or sensitive personal data into event properties **-** Honour cookie and tracking consent **-** Disclose product analytics in your privacy policy **-** Check GDPR (Europe) and DPDP (India) rules for your users, including deletion requests Done right, this is safe and standard. Done carelessly, it is a legal problem. Treat it as moderate legal work, not zero.

How is this different from lead scoring?

**Lead scoring rates marketing signals; PQL scoring rates product usage.** Traditional lead scoring looks at things like job title, company size, email opens and page visits. It works before someone has used your product. PQL scoring looks at what people actually do inside the product: features used, limits hit, teammates invited. It is a much stronger buying signal because it is behaviour, not intent guessed from the outside. Many teams run both. If you have no product usage yet, start with **[lead scoring](/tactics/lead-scoring)** and layer PQLs on once users are active.

How long until this pays off?

**Expect 2-8 weeks to a first working score, and a quarter or two to real lift.** **Rough timeline:** **-** Weeks 1-2: define events and instrument tracking **-** Weeks 3-4: build the score, set a threshold, wire one nudge **-** Weeks 5-8: measure against a control group, cut dead signals **-** Month 3+: conversion lift stabilises and compounds as the model learns The teams that win here ship a rough version fast and tune it, rather than waiting months for a perfect model.

📈 Convert With Data

Telemetry-Driven PQL System

Instrument granular usage events, score each free user on how much value they have shown, and fire automated upgrade prompts, in-app nudges, or a sales handoff at the "aha moment". This replaces broad marketing blasts with usage-triggered conversion driven by what people actually do in your product.

14 min read

Last updated: Jul 24, 2026

💬

"We stopped guessing who to email. Once we scored free users on the three actions that predicted a paid plan, our free-to-paid conversion nearly doubled and our sales team only touched accounts that were already showing intent."

Head of Growth, B2B SaaS

📅 Your First Month

Setup Warm-up Outreach

Week 1

Define activation and aha events

⏱ ~4 hours

Week 2

Instrument event tracking

⏱ ~2-3 days

Week 3

Build a simple PQL score + threshold

⏱ ~1-2 days

Week 4

Wire the first automated nudge

⏱ ~1 day

Ongoing

Measure lift and refine the score

⏱ ~2 hours/week

Week 1

Define activation and aha events

⏱ ~4 hours

Week 2

Instrument event tracking

⏱ ~2-3 days

Week 3

Build a simple PQL score + threshold

⏱ ~1-2 days

Week 4

Wire the first automated nudge

⏱ ~1 day

Ongoing

Measure lift and refine the score

⏱ ~2 hours/week

Is this for you?

Great fit if...

✓You have a free tier, freemium plan, or free trial that people can self-serve
✓Users can reach real value inside the product without a sales call
✓You have enough signups (roughly 100+ new free users a month) to spot patterns
✓You have a developer or two who can add tracking code and call an API
✓Your pricing has a clear upgrade path (usage limits, seats, or premium features)

Try something else if...

You have no free or trial motion and every deal goes through sales first
Your product needs weeks of onboarding before anyone sees value
You get only a handful of signups a month, so there is no data to learn from
You have zero engineering capacity to add and maintain event tracking
Your buyer and your user are completely different people with no in-product overlap

No product usage data yet? Start with lead scoring →

Not the right fit?

If Telemetry-Driven PQL System doesn't match your situation, consider these alternative tactics that achieve similar goals:

Lead Scoring

Prioritize leads based on behavior and demographic signals to focus sales efforts.

advanced•medium•short term

What to expect

2-8 weeks

To first working PQL score

1.5-2x

Free-to-paid lift when tuned well

Ongoing

Compounds as the model learns

Quick math: The first score you ship will be rough. That is fine. The value comes from shipping a simple version, watching which signals actually predict payment, and tightening the threshold every few weeks. Do not wait for a perfect model.

A PQL (product-qualified lead) is a free user whose behaviour shows they are ready to buy. Before you track anything, you need to know which in-product actions actually signal that readiness. The core of this is the aha moment: the first time a user experiences the real value of your product. Everything you build sits on top of getting people to that moment and spotting when they arrive.

The building blocks you are defining here:

🎯 Aha moment: the single action where value clicks (Dropbox: file saved and synced)🎯 Activation: a small set of early actions that predict a user will stick

Find your aha moment

Look at users who converted to paid and ask what they all did that churned users did not:

List the top 5-10 actions a user can take in your product
Pull your last 50 paying customers and note which actions they did in week one
Pull 50 users who signed up and never came back, note what they skipped
The action paying users do and churned users skip is your aha candidate
Write it as one plain sentence: "The user has [done specific action]"

Define 3-5 activation events

These are the milestones on the road to aha. Keep the list short and specific:

Event 1: completed core setup (connected an account, imported data, etc.)
Event 2: performed the primary action once (created a project, sent a message)
Event 3: hit the aha moment (the value action from above)
Event 4: repeated the aha action, showing it was not a fluke
Event 5: invited a teammate or hit a usage limit (expansion signal)

💡 Keep the event list small

Do not track 200 events on day one. Five well-chosen events beat a firehose you will never analyse. You can always add more once the basic score works.

Conversation Flow

A PQL replies to your upgrade email with a question

They are interested. Answer plainly and tie it back to what they were doing.

“Good question. On [Plan name] you would keep everything you have set up and just lose the limit you hit. Happy to walk you through it on a quick call, or you can upgrade from [link] and I will make sure the switch is smooth.”

A user says the nudge felt intrusive

Take it seriously. Acknowledge, and use it to tune your cadence.

“Fair, and thank you for the honesty. We try to only surface that when it is genuinely relevant to how you are using [Product], but we clearly got the timing wrong for you. I have turned it down on your account.”

Sales asks why an account was flagged

Give them the behaviour, not just the number.

“This account crossed the PQL threshold because three users hit the aha action repeatedly this week, one invited two teammates, and they are at 90% of their plan limit. Full activity summary is in the CRM record.”