High-Velocity Performance Creative & Video Ad ProductionPlaybook3 min readUpdated September 2026

Which Signals Actually Mean Someone's Ready to Buy

Every product with a free tier eventually needs to decide which behaviors actually predict someone is ready to pay, and most teams guess at this rather than checking it against their own data. A product-qualified lead definition borrowed from another company's blog post rarely fits your specific product.

This is a guide to building that definition from your own usage data, and to deciding when a signal is strong enough to act on versus when it's just noise.

Finding your real conversion signals, not assumed ones

Start by looking at users who already converted to paid and working backward: what did they do in the free tier before upgrading that non-converters mostly didn't do. This is more reliable than guessing which feature "feels" important, since usage patterns that correlate with conversion are sometimes surprising and rarely match the feature the product team is proudest of.

Look for behaviors that happened before the upgrade decision, not around the same time as it. A signal that only shows up the same week someone upgrades isn't predictive, it's just a symptom of someone who had already decided.

Build your definition in this order:

  1. List the users who upgraded to paid and note what they did in the free tier beforehand that non-converters mostly did not.
  2. Keep only behaviors that happened before the upgrade decision, not ones that appeared alongside it.
  3. Test frequency and depth of use separately, then combine signals instead of relying on login count alone.
  4. Tie the upgrade prompt to a moment when the user hits a real constraint, after they have already gotten value.
  5. Revisit the definition whenever the product changes, and check whether it holds for each customer segment.

Weighing frequency against depth

A user who logs in every day but only ever touches one basic feature is a different kind of engaged than one who logs in less often but explores deeply each time. Depending on your product, either pattern can predict conversion better than the other, so test both rather than assuming frequent login alone means someone is close to paying.

Combine signals rather than relying on one. A user hitting a usage limit and inviting a teammate in the same period is a stronger signal than either behavior alone, since it suggests both a real need and a growing use case within their team.

Deciding when to prompt an upgrade

Prompting too early, before someone has gotten real value, reads as pushy and can sour their impression of the product before they've even decided it's worth paying for. Prompting too late means you're leaving a user who's already convinced sitting on the free tier longer than necessary, with more time to lose momentum or look at a competitor.

Tie the prompt to the moment someone hits a genuine constraint tied to your strongest signal, a usage limit, a feature gate, a team size cap, rather than a generic time-based trigger like "day fourteen of the trial." A constraint-based prompt feels like a natural next step; a calendar-based one feels like a sales tactic.

Getting the messaging right once you've identified the signal

If you're nurturing a product-qualified lead through email before or alongside an in-product prompt, that email still has to get opened to do any good. The average email open rate across industries sits around 35.6 percent1; an upgrade nudge sequence performing well below that is worth testing on subject line alone before assuming the offer itself is the problem.

Reference the actual behavior that triggered the message specifically, not a generic upgrade pitch. Someone who just hit a seat limit responds better to a message about adding teammates than to a broad list of paid-tier features they haven't shown any interest in yet.

Where PQL programs go wrong

The most common mistake is defining a product-qualified lead once and never revisiting it as the product changes. A feature that predicted conversion a year ago may have since become a standard free-tier feature or gotten buried in a redesign, and a stale definition quietly stops predicting anything.

The second is treating every user who crosses the threshold identically. A PQL signal tells you someone is likely ready, not what they specifically need to hear; pair the signal with whatever context you have, company size, use case, stated goal during onboarding, before deciding what the upgrade prompt actually says.

The third is letting the sales team and the self-serve upgrade flow work from different definitions of a qualified lead. If sales is chasing one signal and your in-product prompts are built around another, an account can get two mismatched messages at once, which reads as disorganized to the very users you're trying to convert.

Executive Capability Standard

What Good Looks Like

A good PQL program derives its signal from actual converted-user behavior rather than assumption, combines more than one signal where possible, and ties the upgrade prompt to a genuine constraint the user just hit rather than a generic time-based trigger.

Building The Capability (5-Stage Skill Ladder)

1. Learn:Pull a list of recently converted users and look for usage patterns they shared before upgrading, to build a working hypothesis about what your real PQL signal is.
2. Do Manually:Set a manual threshold based on that hypothesis, and personally review a sample of users who cross it to check whether the signal actually predicts intent to buy before automating anything.
3. Delegate:Have a growth or lifecycle marketer own the PQL definition, refining it against new conversion data and running the upgrade messaging tied to it.
4. Automate:Build the threshold and trigger directly into your product analytics or CRM so upgrade prompts fire automatically the moment a user crosses it, instead of relying on a manual review.
5. Buy:Bring in a product-led growth consultant if you're rebuilding your whole self-serve funnel around this and want an outside review of which signals and prompts actually move conversion.

How to Get Started

Frequently Asked Questions

How do we find our real PQL signals if we don't have a lot of historical conversion data yet?

With limited data, start with a reasonable hypothesis based on which feature most closely maps to your product's core value, and treat it as a working definition to test rather than a final answer. Revisit it as soon as you have enough converted users to check the hypothesis against actual behavior patterns.

Should the same PQL definition apply to every customer segment?

Not necessarily. A signal that predicts conversion well for a small team may not predict it for a larger one with different buying processes and more stakeholders involved. If your segments behave differently enough, it's worth building separate definitions rather than forcing one signal to work across all of them.

Is it better to prompt an upgrade in-product or by email?

In-product prompts tied to the moment someone hits a real constraint tend to convert better because the need is immediate and obvious. Email is more useful for nurturing someone who's shown a weaker signal over time, or for reaching a user who isn't actively in the product at the exact moment they'd benefit from upgrading.

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. Email Open Rate (All Industries). Mailchimp Email Marketing Benchmarks (data updated Dec 2023), 2023.

Related Guides