AI Search Optimization & Generative Engine Optimization (AEO/GEO)Playbook3 min readUpdated September 2026

Scoring Product-Qualified Leads So Sales Gets the Right Ones

Handing sales a list of every active free user and calling them product-qualified leads just relabels your whole free tier as pipeline. A real PQL model needs to separate usage that predicts an upgrade from usage that's just someone getting comfortable with the free plan forever.

Here's how to build that model, hand off leads sales can actually act on, and fix the handoff points where most freemium programs quietly leak conversions.

Which free user actions actually predict an upgrade?

Don't assume total logins or time in product predicts conversion; check your own historical data for which specific actions past upgraders took before they converted, versus which actions free users who never upgraded also took at similar rates. Some usage signals correlate strongly with paid intent, and plenty of common usage doesn't.

Look specifically for actions tied to hitting a real limit of the free plan, like running into a usage cap or trying to access a gated feature, since those tend to predict upgrade intent far better than general engagement metrics that plenty of happy free-forever users also rack up.

For example, list the ten most common actions taken by free users, then compare how often past upgraders took each one before converting against how often free users who never upgraded took it. Actions that appear far more often among upgraders, especially those tied to hitting a plan limit, are candidates for your score. Actions that both groups take at similar rates are just usage and should carry little weight. Repeat this comparison whenever you change the free plan's limits, because the signals that predict upgrades shift when the boundaries move.

How do you score product-qualified leads?

A binary PQL flag hides useful information about how close someone actually is. Build a simple score combining your strongest predictive signals, weighted by how much each one actually correlates with past conversions, so sales can see the difference between a lead just crossing the threshold and one showing strong signals across the board.

Recalibrate that scoring model periodically against actual conversion outcomes. A model built once at launch and never revisited will drift as your product and user base change, and a stale model quietly starts sending sales worse leads without anyone noticing until pipeline quality slips.

Hand off with context, not just a score

A lead that shows up in sales' queue with only a numeric score and no context forces the rep to guess what to say. Pass along the specific triggering action, such as hitting a usage limit or exporting data repeatedly, so the rep's outreach can reference something concrete and relevant instead of a generic upgrade pitch.

That specific context is also what separates a helpful, well-timed outreach from one that feels like generic sales spam to someone who was in the middle of getting real value from your free plan.

A product-qualified lead handoff to sales should include:

  • The specific triggering action, such as hitting a usage limit or exporting data repeatedly.
  • The PQL score, so the rep can see how far past the threshold the lead is.
  • The signals behind that score, weighted by how strongly they predicted past upgrades.
  • Whether the user has already ignored outreach, which suggests a self-serve upgrade message instead of another call.

Give the upgrade page a fair test of its own

Compare your in-product upgrade page against the broader median: landing pages convert at 6.6% across all industries1, and a page shown to someone who already hit a real limit and clicked through should usually do better than that baseline, not worse.

Test the upgrade page's copy and pricing presentation the same way you'd test any other conversion page, rather than assuming a user who's already engaged with your product doesn't need a well-built page to complete the upgrade.

Watch for sales burning out a legitimately engaged free user

Not every PQL wants a sales call; some just want to self-serve an upgrade in-product without talking to anyone. Offer a clear self-serve upgrade path alongside any sales outreach, and let usage patterns, like a user who's never responded to outreach but keeps hitting limits, guide you toward self-serve messaging instead of another call attempt.

Track how often PQLs convert self-serve versus through a sales touch, and let that ratio inform how much sales capacity you actually need pointed at this motion, rather than assuming every product-qualified lead needs a human conversation to close.

Revisit that ratio whenever you change the self-serve upgrade flow itself. A meaningful shift toward self-serve after a redesign is a sign the earlier ratio was reflecting a clunky flow rather than genuine buyer preference for talking to a person, and it's worth flagging that finding directly to whoever owns the product experience so the win gets attributed correctly rather than credited only to sales enablement.

Executive Capability Standard

What Good Looks Like

A working PQL program scores leads on actions that actually correlate with past conversions in that product's own data, gets recalibrated periodically against real outcomes, hands sales the specific triggering context rather than a bare score, and offers a genuine self-serve upgrade path alongside outreach.

Building The Capability (5-Stage Skill Ladder)

1. Learn:Pull your last several quarters of upgrade data and compare which actions past upgraders took most often against actions free users who never upgraded also took at similar rates.
2. Do Manually:Build a simple weighted score in a spreadsheet from your strongest signals and manually review a sample of flagged leads with sales before automating anything.
3. Delegate:Have a growth or RevOps owner maintain the scoring model and recalibrate it against conversion data each quarter, reporting drift back to you.
4. Automate:Wire the scoring model directly into your CRM so qualified leads route to sales automatically with the specific triggering action attached, not just a numeric score.
5. Buy:Bring in a product-led growth consultant if your usage data is too thin or noisy to build a reliable model without outside help setting up the initial analysis.

How to Get Started

Frequently Asked Questions

Should every active free user count as a product-qualified lead?

No. Total activity alone doesn't predict upgrades well. Build a score from the specific actions that correlate with past conversions in your own data, particularly ones tied to hitting a real limit of the free plan, rather than treating general engagement as qualification.

How often should a PQL scoring model be recalibrated?

Periodically, checked against actual conversion outcomes rather than left as is from launch. A model that isn't revisited will drift as your product and user base evolve, and a stale model can quietly send sales weaker leads without anyone noticing right away.

Do all product-qualified leads need a sales call?

No. Some engaged free users prefer to upgrade themselves in-product without a conversation. Offer a clear self-serve path alongside sales outreach, and track how often PQLs convert each way, so sales capacity goes where it's actually needed.

Sources

Where we quote a benchmark, we show its source. Other figures in this guide are estimates or general guidance, so check them against your own numbers.

  1. Landing Page Conversion Rate (Median, All Industries). Unbounce Conversion Benchmark Report.

Related Guides