Cometly
AcademyModule 03 · Product-Led Growth Reports
PLGStrategyLesson 3.5·6 min read

Optimizing Meta and Google for paying customers, not trials

Same ad spend, dramatically different customer mix.

The single biggest paid-acquisition unlock for PLG SaaS is switching Meta and Google’s optimization event from Trial Started to New Customer. The same ad spend produces a dramatically different customer mix once the platforms are hunting for paying users instead of free signups.

Why this matters

The platforms’ algorithms are extremely effective at finding more of whatever you tell them is success. When that’s 'free trial signup,' they find users who love free things — including users who will never pay. When it’s 'New Customer with $X value,' they find users who buy.

Section 01

Volume threshold

Meta and Google both need at least 50 conversions per week to optimize reliably on a deeper-funnel event. If your weekly New Customer volume is below that, optimize on the closest deeper event you do have volume for (often Trial Converted) and use New Customer as a secondary signal.

Pair the New Customer event with a value parameter — typically the first month’s MRR or expected LTV — so the platform can run value-based bidding instead of pure conversion bidding.

  • Above 50 weekly New Customers → optimize on New Customer with value bidding
  • 20–50 weekly Trial Converted → optimize on Trial Converted, push New Customer for retargeting
  • Below 20 weekly Trial Converted → optimize on Trial Started, push deeper events for audience enrichment
Section 02

Expected patterns

When you switch the optimization event from Trial Started to New Customer, expect cost-per-trial to roughly double and trial volume to drop 30–60%. That’s the algorithm doing its job — finding higher-intent users.

The number that should improve is cost-per-customer or LTV ROAS. If you don’t see improvement in 6–8 weeks, the issue is usually creative-audience fit, not the optimization event.

Common pitfalls

What to watch for.

  • Switching events without enough volume

    Below 50 weekly conversions the algorithm can’t learn. Stay on the higher-volume event.

  • Cutting the trial event entirely

    Keep the trial event live as a secondary signal for retargeting and lookalikes. Just don’t optimize on it.

  • Not sending value with the New Customer event

    Without a value parameter, the platforms can’t do value-based bidding. Always pass first-month MRR or expected LTV.

Key takeaways

Recap.

  • Push New Customer events to Meta, Google, LinkedIn, TikTok, and Bing simultaneously
  • Wait for at least 50 New Customer events per week before switching the optimization event
  • Pair the New Customer optimization event with a value parameter to enable ROAS bidding
  • Expect CPL and trial volume to drop and paying-customer volume to rise — that’s the goal
  • Keep the trial event live as a secondary signal for retargeting and lookalike audiences
Put it into practice

Build this report inside your own Cometly workspace.

Most lessons can be wired up in a single 30-minute onboarding call. Connect your stack live and walk away with a working dashboard.