Four levels of PLG optimization
On optimizing top pages, sequences, loops, and ecosystems.
11 years scaling Dropbox and Asana taught me a lot about PLG.
Since leaving Asana, I’ve been sharing insights with VCs, growth leaders, and consulting clients, and have found that many of my mental models resonate. So I’m starting to share them on LinkedIn and Substack. First up: levels of PLG optimization.
PLG optimization moves through four discrete levels, each requiring distinct approaches:
Level 1: Optimize “Golden” Pages
Complexity: 1/10
Improve conversion rate (CRO) on standard, high-converting pages: homepage, pricing, signup, contact sales, in-app payments/billing
Requires simply awareness this can be done and willingness to try
Process is to implement best practices easily available via Google or advisors
💡 Low effort drives big impact, but most underestimate when they’re small ($XM ARR)
Level 2: Optimize Standard Flows
Complexity: 3/10
Optimize standard multi-screen sequences: onboarding, invite, churn, signup/trial emails
Requires an experienced growth PM + design, eng, in-prod A/B testing
Process is to develop hypotheses, prioritize initiatives on estimated effort vs. impact, then test & iterate
💡 Game-changing impact, especially to early adoption and user growth
Level 3: Optimize Proprietary Loops
Complexity: 6/10
Optimize unique behavior loops that drive *your* new user activation, high-value feature usage, account expansion, and monetization
Requires PMs, strong data science and customer insights, then great XFN collaboration to optimize core product surfaces
Process is to develop deep customer insight, identify shorthand adoption/value creation/monetization metrics, then optimize. Popular example is Facebook’s "7 friends in 10 days": FB had to study users, identify a simple yet moveable metric, then test interventions to inflect
💡 L3 is a high effort, defensible moat: copying L1/L2 is easy, L3+ is harder to reverse-engineer
Level 4: Tune the Ecosystem
Complexity: 10/10
Tune the outputs of the entire PLG ecosystem: topline metrics (DAU, MAU, customer retention, NDRR) are indicators of how a PLG ecosystem is working as a whole, and nearly impossible to optimize directly
Requires an XFN team that understands users, develops intuition about how the overall ecosystem functions, and is open to taking “big swing” risks
Process is to develop intuition on how complex loops intersect, then test ecosystem nudges to drive improvement. For example, at Asana we figured out when and how product UI simplification could help users focus to discover & use higher-value features, improving monetization over time
💡 L4 upside is nearly unlimited, particularly when including PLG to Sales (PLS) loop intersection