A9 · Research synthesis · Article schema
PLG Benchmarks: What the Data Actually Shows
The bottom line
Sources reviewed
| Source | Finding | Quality | Notes |
|---|---|---|---|
| OpenView Partners, PLG Benchmarks Report 2022-2024 | Median free-to-paid conversion: 3-5% overall; 8-15% for bottoms-up enterprise tools (Slack, Notion tier). Median time-to-activation (first meaningful action): 3-7 days. Product-qualified lead (PQL) to close rates: 15-25% | directional | OpenView invests in PLG companies; sample is skewed toward high-quality PLG businesses in their network. Figures should be treated as best-practice benchmarks, not median-market estimates |
| Slack S-1 and investor letters, 2019-2023 | At IPO, ~30% of Slack's paid organizations began as free teams. Daily active user to paid conversion for teams with 3+ active members was approximately 30-35%. NRR was 143% in FY2019, declining to 125% as base grew | established | Public filing data — highest reliability. Slack's conversion rate is an outlier due to strong network effects; not a representative PLG benchmark |
| Figma S-1 prep materials / Bloomberg reported acquisition metrics, 2022 | At time of Adobe acquisition announcement, Figma had ~$400M ARR growing 100%+ YoY with a user base of 4M+ active users; viral coefficient estimated at 0.6-1.0 (each user inviting 0.6-1.0 new users on average) | directional | Figma was private; exact metrics come from press reports citing sources familiar with the S-1 prep. Direction and order of magnitude are well-reported but not verifiable against public filings |
| Bessemer Venture Partners, State of the Cloud 2023 | PLG companies in the BVP Nasdaq Cloud Index (n=74) averaged NRR of 120% in 2022, declining to 108% in 2023. Top quartile PLG companies sustained NRR 130%+ through the downturn | established | Public company dataset — high reliability for direction and magnitude |
| Reforge / Brian Balfour, Growth Frameworks 2020-2022 | Activation rates (users who complete the 'aha moment' action) average 20-40% for consumer apps; 10-25% for B2B SaaS; companies with activation > 40% have retention curves that flatten 2x higher than those with activation < 20% | directional | Practitioner-compiled; sample is Reforge community members (above-average execution). Direction is consistent with first-principles; exact figures are context-dependent |
The mechanism
PLG growth works through two compounding loops: the product loop (users invite collaborators or share outputs, triggering new signups) and the data loop (usage generates insights that improve the product, raising retention and activation). The viral coefficient K measures the product loop: K = invitations per user × acceptance rate. K > 1 produces exponential organic growth; K = 0.5-0.9 (as in most B2B PLG tools) produces meaningful but sub-viral organic contribution that reduces paid CAC by 20-50% relative to purely sales-led motions.
Free-to-paid conversion is highly dependent on the product's natural team size and collaboration radius. Single-player tools (analytics, writing) convert at 3-5%. Multi-player tools where the product creates switching costs as team size grows (Figma, Notion, Slack) convert at 15-30% of active teams. Enterprise PLG (adding sales for >50-seat or $100k+ ACV accounts) layers sales-assisted expansion on top of bottoms-up adoption, achieving NRR > 120% by combining product expansion with sales-negotiated multi-year contracts.
Implications for GTM operators
PLG leaders should instrument activation obsessively — the 'aha moment' is the single highest-leverage conversion point. Benchmarks from vendor reports should be adjusted downward by 30-40% to reflect survivorship bias in survey respondents. Viral coefficient should be measured as a lagging indicator of product-loop health, not as a vanity metric; K declining below 0.3 is an early warning of loop degradation. Enterprise PLG should add sales capacity only after product-led expansion is demonstrably working (NRR > 110% from product-driven expansion alone).
What this doesn't settle
This synthesis rates the confidence of the directional finding as directional, not proven. The sources reviewed range from public company filings (highest reliability) to vendor-reported survey data (lower reliability, potential commercial interest). Exact percentages and benchmarks should be treated as order-of-magnitude estimates rather than precise universal figures.
The synthesis does not settle: (a) whether findings from the reviewed sources generalise to your specific market segment, company stage, or ACV tier; (b) causal direction where only correlational data is available; (c) how findings will shift as market conditions evolve post-2024. Practitioners should treat this as prior-setting evidence that warrants in-house measurement, not as a substitute for first-party data.
Related concepts
How to cite
@misc{shalvi_gtm_synthesis_plg_benchmarks_2026,
author = {Singh, Shalvi},
title = {PLG Benchmarks: What the Data Actually Shows},
year = {2026},
url = {https://shalvisingh.com/gtm/syntheses/plg-benchmarks},
note = {GTM World Model A9 Research Synthesis}
} Singh, S. (2026). *PLG Benchmarks: What the Data Actually Shows*. GTM World Model. https://shalvisingh.com/gtm/syntheses/plg-benchmarks