PLG Removes Human Work from Growth
Blake Bartlett, partner at OpenView, on the future of product-led growth
PLG works best when the product turns work that used to require people into software, which means growth spending shifts from hiring more reps and support staff to improving onboarding, pricing, and in product prompts. The real test is not a special PLG metric, but whether self serve usage produces better CAC payback, strong retention, and clear funnel health from signup to activation to paid conversion, even before a big sales team arrives.
-
In practice, de laboring means users can discover the product, sign up, reach value, and often pay by card without talking to anyone. That is why PLG companies track classic SaaS outcomes like ARR growth, net retention, and CAC payback alongside consumer style funnel steps like activation, 7 day retention, 14 day retention, and conversion to paid.
-
The best PLG examples automate different parts of growth. Calendly turns every meeting link into distribution and reached 20 million users with minimal capital before layering enterprise sales. Sentry gets leverage from an SDK that installs into code, captures usage automatically, and drove 70% of revenue through self serve while keeping sales and marketing headcount lean.
-
PLG rarely removes humans forever. It changes where they show up. Airtable watches product signals like shared views, seat growth, and usage of integrations to know when to step in with sales or services. DocSend found that support, live chat, and better packaging converted more customers than a large outbound motion, because helpful guidance amplified a self serve funnel instead of replacing it.
Going forward, the strongest PLG companies will look less like pure freemium funnels and more like automated growth systems with selective human assist. The winners will use product telemetry to spot when a user or team hits a ceiling, then add support, sales, or AI at that moment, which keeps the engine software driven while still expanding into larger accounts.