A product-qualified lead (PQL) is an account whose behavior inside your product shows it’s likely to pay, and whose firmographics make it worth a sales conversation. Define it from the few usage behaviors that separate trials that converted from trials that didn’t, add an ICP fit filter, and put it in front of a rep while the trial is still active. A form fill says someone is curious; usage says they’re getting value.
PQL vs MQL vs lead score: what changes
In a sales-led funnel, marketing qualifies leads on engagement: pages visited, content downloaded, demos requested. In a product-led funnel the prospect is already using the product, so those signals are weak next to what the account actually does there.
| MQL | Lead score | PQL | |
|---|---|---|---|
| Based on | Marketing engagement | Weighted mix of fit and engagement | Product usage plus account fit |
| Typical unit | Contact | Contact | Account or workspace |
| Trigger | Form fill or content download | Points crossing a threshold | Reaching defined usage milestones |
| What it shows | Interest | Depends on the model | Experienced value |
| Timing pressure | Low | Low | High: the trial ends or free usage stalls |
Three things change when you move to PQLs:
- The unit moves from contact to account. In a workspace product, several users shape the buying decision.
- The data source moves from your marketing automation platform to your product database or analytics tool.
- The clock starts. A PQL in a 14-day trial is worth far more on day 5 than on day 13.
A PQL definition doesn’t have to replace your lead score. Many teams keep a fit-and-engagement score for inbound demo requests and run PQLs alongside it for self-serve signups. If you already have a lead scoring model in HubSpot, product milestones can become its strongest inputs. If you’re still deciding whether sales should touch self-serve signups at all, settle product-led vs sales-led growth first.
Finding the usage behaviors that predict conversion
Don’t pick PQL criteria on a whiteboard. Pull them from your own conversion data.
- Export recent trial or free accounts that had a full chance to convert, flagged by whether each became paying. With fewer than a hundred conversions, treat the output as a hypothesis to test.
- List 15 to 30 candidate behaviors that plausibly signal value: invited a teammate, connected an integration, imported data, used a core feature repeatedly, returned on three separate days, hit a plan limit, viewed the billing page.
- Compare conversion rates for accounts that did each behavior within the first few days against those that didn’t. You want a large gap and enough volume to matter.
- Check timing. A behavior that predicts conversion but usually happens on the day of purchase is useless for routing. You need signals early enough to act on.
- Combine two to four behaviors into one definition, such as “invited at least two teammates and connected an integration within seven days.”
Here’s what the output looks like, using made-up numbers:
| Behavior in first 7 days | Share of trials that did it | Converted if yes | Converted if no |
|---|---|---|---|
| Invited 2+ teammates | 25% | 30% | 6% |
| Connected an integration | 40% | 20% | 7% |
| Created 5+ projects | 15% | 18% | 11% |
| Viewed billing page | 10% | 35% | 9% |
In this hypothetical data, teammate invites and integrations are strong and common enough to route on. Project count shows a weaker gap. The billing page view is strong but often happens late, and sometimes it’s just a price check, so use it as a secondary trigger rather than the core definition.
Correlation isn’t causation. Teams that invite colleagues may convert because they were larger companies to begin with. That’s fine for routing, but don’t assume forcing invites in onboarding will lift conversion by the same margin.
Combining product signals with firmographic fit
Usage alone creates false positives. A student or a solo freelancer can be your most engaged user and never be worth a rep’s time. Fit alone takes you back to MQLs. The PQL is the overlap.
Score fit from firmographic data: company size, industry, region, tech stack and whether the signup used a business email domain. Enrichment tools can fill much of this at signup. Base the fit criteria on your closed-won deals, not on the customers you wish you had.
Then sort accounts into a simple grid:
| Strong usage | Weak usage | |
|---|---|---|
| Strong fit | PQL: route to sales now | Stuck: offer setup help from a rep or CSM |
| Weak fit | Self-serve: in-app upgrade prompts and email | Automated nurture only |
The “stuck” box deserves its own play: good-fit accounts that stalled often respond to a short, helpful note about setup. Track it separately from PQLs so it doesn’t blur your conversion numbers.
Add a deal size floor too. If an account’s likely plan and seat count would bring in less than a rep’s time costs, let self-serve handle it.
Getting product data into the CRM
Reps work in the CRM, so the PQL has to appear there with enough context to act on. There are three usual paths:
- Direct integration or API. Your app writes key properties to the CRM company record. Simple, but engineering owns every change.
- Customer data platform. A tool like Segment sends product events to both analytics and the CRM.
- Reverse ETL from the warehouse. You compute PQL status in SQL in your data warehouse, and a tool like Hightouch or Census syncs it to HubSpot or Salesforce. Best when the definition joins usage with billing and enrichment data.
Whichever path you choose, sync account-level summaries, not raw events. A rep needs “three active users, integration connected, nine days left in trial,” not thousands of clicks on the timeline.
Minimum fields to sync:
- Workspace or account ID, mapped to the CRM company record
- Signup date and trial end date
- Current plan and seat count
- Active users in the last 7 days
- Each PQL behavior as a date field (blank if not reached)
- PQL status and the date it was reached
- Fit tier from enrichment
Sync at least daily, and hourly for short trials. Match workspaces to companies carefully: personal email signups, several workspaces from one company and users from multiple domains all break naive matching, and a bad match sends the PQL to the wrong rep.
Routing rules and sales plays for PQLs
When PQL status flips, three things should happen automatically: the account gets an owner, the owner gets notified with context, and a task is created with a deadline.
Routing rules
- Give PQLs their own queue so they aren’t buried under MQLs, then assign by territory, segment or round robin as you do for other inbound.
- Set a response time in your sales SLA. Same business day is a reasonable target when a week or two of trial remains.
- If the account already has an owner, such as an open opportunity or an existing customer’s new workspace, route to that owner.
- Once a rep engages, suppress the automated upgrade emails that would contradict the conversation.
Sales plays by trigger
One script for every PQL wastes the signal. The trigger tells the rep what the account needs.
| PQL trigger | What it suggests | Opening play |
|---|---|---|
| Teammate invites | Evaluating for a team | Offer rollout help; discuss team plan and admin needs |
| Plan or usage limit hit | Getting value, now blocked | Explain options for the limit; quote the right tier |
| Integration connected | Moving toward production use | Offer technical help, security review or migration support |
| Billing page viewed after activation | Checking price | Answer pricing questions; discuss annual terms or procurement |
Outreach should read as help, not a pitch. “Your team connected Salesforce and has five people in the workspace. Teams at this point usually ask about SSO and admin roles; happy to walk you through both” works. “Just checking in on your trial” doesn’t.
Measuring PQL conversion and refining the definition
Review these monthly:
- PQL rate: the share of new signups that become PQLs. Too high and the bar is too low; too low and reps starve.
- PQL-to-paid conversion compared with non-PQL signups. If the gap is small, the definition isn’t predicting anything.
- Time from PQL to first touch. Slow response wastes the signal.
- Sales-assisted vs self-serve conversion among PQLs, with average deal size for each.
- Revenue from PQL-sourced deals, so the program is judged on money, not volume.
The hard question is whether sales adds value or just claims deals that would have closed anyway. The cleanest answer is a holdout: randomly leave a small share of PQLs without outreach for a few weeks, then compare conversion rate and deal size. If the contacted group doesn’t convert better or buy bigger, rework the plays or raise the fit threshold.
Let reps mark a PQL as bad with a reason (wrong fit, student, competitor). Those reasons are your best input for tightening the definition. Revisit it quarterly and after major product or onboarding changes.
Common mistakes
- Defining PQLs at contact level. One engaged user in a ten-person workspace isn’t a team adopting the product.
- Using signup as the trigger. A signup is a lead, not a PQL.
- Ignoring fit. Reps burn hours on students, competitors and accounts too small to pay for the conversation.
- Routing too late. A definition that fires on day 12 of a 14-day trial leaves no time to help.
- Colliding messages. The prospect gets a personal note from a rep and an automated “your trial is ending” blast on the same afternoon.
PQL routing sits where lifecycle marketing meets sales operations, and it’s part of how I build lifecycle and CRM systems for product-led companies.
Get it built
If your trials convert but sales is chasing the wrong accounts, or none at all, I can define the PQL model, wire it into HubSpot or Salesforce and set up the plays. Start with a Growth Audit, $1,500 fixed and credited if we continue. See pricing or get in touch.