Engineering the funnel around product-qualified leads

The vocabulary: MQLMQLA Marketing Qualified Lead (MQL) is a prospect whose engagement and fit signals indicate they are more likely to become a customer, justifying handoff toward sales.View full definition →, PQL, SQLSQLSales Qualified Lead: a prospect the sales team has validated as ready for direct outreach and a proposal, having passed clear qualification criteria.View full definition →

Let us define the three lead types before we build anything.

MQL (Marketing Qualified Lead): someone who has shown interest through marketing activity. They downloaded an ebook, attended a webinar, or hit a pricing page three times. Interest, not usage.

PQL (Product Qualified Lead): someone who has used the product (usually a free trial or free tier) and hit behaviors that predict they will buy. Usage, not just interest.

SQL (Sales Qualified Lead): a lead a salesperson has vetted and accepted as worth active pursuit. This is the handoff to a human closer.

The old B2B software playbook ran MQLMQLA Marketing Qualified Lead (MQL) is a prospect whose engagement and fit signals indicate they are more likely to become a customer, justifying handoff toward sales.View full definition → to SQLSQLSales Qualified Lead: a prospect the sales team has validated as ready for direct outreach and a proposal, having passed clear qualification criteria.View full definition →: capture an email, nurture with content, pass to sales. That works when buying happens before using.

In modern SaaS, especially product-led growth (PLG) companies where the product itself drives acquisition, people use before they buy. Slack, Figma, Notion, and Calendly all let you get value before a contract. The PQL is the signal that fits that motion.

Why PQLs convert better

The logic is simple: intent shown through action beats intent shown through clicks.

Someone who invited five teammates and automated a workflow has already integrated your product into their day. Someone who downloaded a PDF has integrated nothing.

Industry practitioners consistently report that PQLs convert to paid at meaningfully higher rates than MQLs, though the exact multiple varies by product and is often cited loosely. Treat any specific figure you see as an estimate, not gospel. The directional truth is solid: usage predicts revenue better than form fills.

Building a PQL scoring model

A PQL score is a number that ranks trial users by how likely they are to buy. You build it from activation signals, the in-product events that correlate with becoming a paying customer.

Step 1: Find your activation signals

Sit down with product analytics and answer one question: what did users who converted do that users who churned did not?

Common signal categories:

• Breadth signals: number of features used, number of integrations connected.
• Team signals: seats invited, teammates who accepted, shared workspaces created.
• Depth signals: hitting the "aha moment," the point where the user first experiences core value. For a design tool it might be publishing a first prototype. For an email tool, sending a first campaign.
• Frequency signals: days active in the first week, logins in first 14 days.
• Fit signals: company size, work email domain (not gmail.com), job title from enrichment.

Step 2: Weight the signals

Not every signal is equal. Seats invited usually predicts more revenue than a single login because SaaS pricing is often per seat and adoption spreads through teams.

Here is a simplified scoring model expressed in plain pseudocode:

The weights are not magic. You set them, then you validate them against who actually converted, then you adjust. Start rough. A model that flags obvious buyers beats no model.

Step 3: Set a threshold

Pick a score that separates "sales should reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.View full definition → out" from "keep nurturing in-product." Set it too low and you drown reps in weak leads. Set it too high and you miss real buyers.

A practical starting method: look at your last 90 days of conversions, find the score most of them crossed, and set the threshold just below it.

For a deeper walk through activation metrics, Amplitude's Product Analytics Playbook is a solid free resource on defining aha moments and activation events.

Mapping the handoffs

A score is useless if nobody acts on it. The hard part of PQL programs is organizational: three teams have to hand off cleanly.

Marketing owns the top

Marketing drives sign-ups and free-tier activation. In a PLG motion, marketing's job shifts from "generate demo requests" to "generate qualified trials and help users activate."

That means marketing cares about in-product onboarding, lifecycle emails, and activation rate, not just clicks. A marketer running a Google Ad now needs to know whether that traffic activates, not just whether it converts to a trial.

Product owns the middle

Product surfaces the signals and drives activation inside the app. Product tells you when the aha moment happened. Product runs the in-app nudges ("Invite your team to unlock shared boards").

This is the collaboration that breaks most often. Marketing and product historically did not share dashboards. In a PQL model they must share one definition of activation.

Sales owns the qualified handoff

When a user crosses the PQL threshold, a signal fires to sales. A rep sees context: this account invited six people, connected Salesforce, and is on a free plan. The rep reaches out with relevance, not a cold script.

Note the mindset shift. Sales is not chasing strangers. They are talking to people already getting value, often to expand an account or move a team from free to paid. This is sometimes called a "sales-assist" motion layered on top of PLG.

The handoff mechanics

The plumbing matters. A typical stack:

1. Product analytics tool (for example Amplitude or Mixpanel) tracks events.

2. Events feed a customer data platformcustomer data platformA Customer Data Platform unifies customer data from all sources into persistent, actionable profiles that other systems can use.View full definition → or reverse ETLETLETL (Extract, Transform, Load) is a data integration process that pulls data from sources, reshapes it into a consistent format, and writes it into a target system.View full definition → tool.

3. The PQL score lands in your CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition → (for example Salesforce or HubSpot) as a field on the account.

4. A workflow alerts the assigned rep and creates a task.

Speed counts. A PQL is hottest right after the activation spike. A reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.View full definition →-out days later hits a colder user.

Common failure modes

Scoring on vanity signals. Logins alone are weak. A user can log in daily and get no value. Weight signals that tie to your core value, not just presence.

Ignoring account-level signals. In B2B SaaS the buyer is often a team, not an individual. Roll up individual signals to the account. Five users each hitting the aha moment inside one company is a far stronger buying signal than one power user.

No feedback loop. Sales must tell marketing and product which PQLs were actually good. Without that loop your weights never improve. Set a monthly review where sales grades a sample of PQLs.

Treating PQL as a rename of MQL. If you just relabel your old leads, you gained nothing. The whole point is that the signal comes from real product usage.

Firing sales at self-serve buyers. Some users want to buy alone with a credit card. If a rep interrupts a smooth self-serve purchase, you can create friction. Segment: let low-touch users convert themselves, route high-value team accounts to sales.

A quick worked example

Imagine a SaaS scheduling tool with a 14-day free trial.

Marketing runs a content campaign that drives 1,000 trial sign-ups. Product data shows 300 reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.View full definition → the aha moment (booking their first meeting through the tool). Of those, 90 invite at least two teammates and connect a calendar.

Those 90 cross the PQL threshold. Instead of a rep cold-calling all 1,000 sign-ups, sales gets 90 warm, contextualized leads. The other 710 stay in automated lifecycle nurture, and many will self-serve or activate later.

That is the efficiency gain: sales time goes to accounts already showing they will pay.

Key takeaways

• PQLs are leads defined by product usage, not marketing engagement. They convert better because action predicts intent more reliably than clicks.
• Build the score from validated activation signals: seats invited, aha moment reached, breadth of use, and firmographic fit. Weight them, then validate against real conversions and adjust.
• The model lives or dies on handoffs. Marketing drives qualified trials and activation, product surfaces the signals, and sales works the accounts that cross the threshold.
• Roll signals up to the account level in B2B, and act fast while the lead is hot.
• Close the feedback loop: have sales grade PQLs monthly so your weights keep improving. A rough model that ships beats a perfect model that never launches.