60% and 40/40/20: A Trial-to-Paid Conversion Model

Priya Raman’s two-part model ties trial-to-paid conversion to a 60% probability threshold and gives every team one owned metric. At Crazy Egg, its growth team devotes 40% of its time to qualified visitors, 40% to trial signups and 20% to engaged trials.

Raman draws on 15 years at companies including Crazy Egg and Help Scout. She says the model can cover self-service and complex, sales-led businesses; organizations with 500 clients, 10,000 clients or 25 million users; and marketing, compliance, support, analytics and luxury retail. Those cases define the model’s reach, not its guaranteed results.

“The key thing is to look at the individual levers that move someone from Engaged to someone who has a high probability of converting,” Raman said. For her, those levers tip a user above a 60% probability of becoming a paying customer. The 60% figure is her working threshold, not a published benchmark.

What should teams measure before optimizing conversion?

Begin with a wide set of funnel metrics rather than a company-wide north star, which can hide where customers stall. Raman said, “I never look at a sole Trial metric.” Her first metric groups are specific:

• Qualified visitors: Visitors who reach pages that most strongly correlate with trial signup. Top-of-funnel teams should treat this as a core metric and ignore traffic that does not fit that pattern.

• Trial signups: People who register for the product.

• Onboarded trials: People who complete setup and can use the product.

• Engaged trials: People who sign up, complete setup and return to take an action.

• Highly correlating levers: Behaviors most associated with movement from an engaged trial to a high probability of paid conversion. Teams should use these signals to decide where to intervene.

Company examples: At Help Scout, a certain number of customer conversations plus adding one additional user correlated with paid conversion; at LiveIntent, speed in spending a fixed budget amount against a goal; at Chartbeat, time in the dashboard during the first seven days.

Crazy Egg has identified its own pattern, but Raman cannot share it yet. That missing detail matters: the method requires company-specific behavioral evidence rather than a universal checklist.

At one company, leaders rejected the methodology and would not let Raman hire a data scientist, so she hired a consultant. The analysis took 20 hours of work, and Raman says the resulting data transformed the business.

• New customers: Trial users inside a critical adoption window; failure to adopt raises the chance of churn during the contract or after it. Raman uses trial plus 30 days as the default, Help Scout’s critical window was 45 days, and Crazy Egg’s was 60 days.

Teams should treat new customers separately during the trial and their first 30 days as customers, providing what Raman calls training wheels for successful product use. Together, these stages and levers materially affect trial-to-paid conversion.

Trial segmentation comes last. At Help Scout, incoming customers varied by support-team size, and the highly correlating levers differed for teams of three, 10 and 25 people. Raman advises teams to establish the broad metrics first, watch behavioral cohorts—groups of users with similar patterns—cluster naturally, then retain or reset the clusters; she identifies pricing as a way to reset them. She says teams repeatedly have to work out this segmentation, even when they join or enter a company.

How should teams own the metrics?

Raman rejects one North Star Metric—a single company-wide measure—for the entire company. She prefers simple formulas that give individual metrics context and make them harder to game. At LiveIntent, she created one formula that unified buyers and sellers in the marketplace and produced a three-step view of business health and where to focus.

Every discrete team should have its own north star metric and enough capacity to execute against it. Each team owns one primary metric and shares responsibility for a secondary metric, which becomes the receiving team’s primary metric. At Crazy Egg, the Trial Signups team also watches Engaged Trials because Engaged Trials belongs to the next team in the customer journey.

What if the team is small?

When a company lacks separate teams, Raman recommends compartmentalizing the work in 10% increments, equivalent to half or full business days. Crazy Egg’s lean Growth Team splits its attention across three metrics:

• 40% — Qualified visitors: Monday and Thursday.
• 40% — Trial signups: Tuesday and Friday.
• 20% — Engaged trials: Wednesday.

When does this model fail?

It depends on more than a dashboard. A single company-wide north star can conceal stage-level failure; unqualified segmentation can impose categories before the data forms them; and weak ownership leaves the next team without a clear handoff.

The examples are correlations, so teams still need to test whether changing a behavior causes conversion. Raman’s inability to disclose Crazy Egg’s pattern also shows why the 60% threshold cannot substitute for company-specific analysis.

Raman’s model points toward product growth as a chain of measured handoffs: each stage has a metric, an owner and a time box, supporting sustainable conversion before the next inflection.