The PLG Playbook Is Being Rewritten in 2026

What the original playbook assumed

Classic PLG was built on three bets: that low-friction self-serve was a meaningful differentiator, that product usage reliably signalled purchase intent, and that viral loops would keep customer acquisition costs structurally low. In 2021 and 2022 those bets paid off. In 2026 all three are under pressure at the same time.

Self-serve onboarding is no longer a differentiator. It is the floor. Every serious B2B SaaS product offers it. Which means a user signing up and activating tells you much less than it used to. The signal has not disappeared entirely, but product-qualified leads are less predictive of revenue than they were three years ago, and the teams still optimising hard for PQL volume are learning this the slow way.

I often see this gap most clearly when a team celebrates activation metrics while the sales team quietly reports that the deals they are closing look nothing like the users driving those numbers.

What AI changed, and what it did not

AI has made it dramatically cheaper to build features. The feasibility barrier has collapsed. Which means competitive pressure has shifted almost entirely onto whether a feature creates real value, whether the business case for it holds, and whether users can actually succeed with it. The three dimensions PLG metrics have always been worst at measuring.

This is the uncomfortable version of the AI story in product. Not "AI helps us move faster" but "moving faster no longer protects you." The teams I work with who are rewriting their PLG strategy in 2026 have noticed this. They are investing more in understanding expansion behaviour from week one, not retrofitting expansion into a freemium architecture that was only ever designed for acquisition.

If you are still instrumenting primarily for top-of-funnel activation, you are measuring the part of the model that has lost the most explanatory power.

The acquisition user and the expansion buyer are not the same person

Here is the assumption that causes the most structural damage and gets discussed the least: most PLG products were built assuming the person who signs up is the same person who decides to expand. In practice, they rarely are.

The individual contributor who starts a free trial is not the procurement lead who approves the enterprise contract. The engineer who activates your developer tool is not the VP who decides whether to standardise on it. Classic PLG optimised the experience for the first person while assuming the second person would follow. That worked when buying decisions were smaller and faster. It works less well when enterprise buyers are entering PLG funnels earlier and converting through sales-assisted paths more often.

This is not an argument that every product needs a full enterprise sales motion. It is an argument that PLG and sales-led are not two separate strategies to choose between. They are one integrated model, and the seam between them is where most PLG teams are losing revenue they do not realise they are losing.

Making this choice explicitly, as an architectural decision, is a different conversation than asking which features to build next. It requires a CPO to ask: who is acquiring, who is expanding, and have we designed the product and the go-to-market for both, or only for one?

What this means at two levels

For a CPO, the question is whether your growth model has been revisited since the assumptions it was built on were still true. Not the tactics. The assumptions. If the last honest conversation about the architecture of your PLG motion happened before 2024, it is probably worth having it again, with the current data in the room rather than the data that validated the original bets.

For a PM in the team, this surfaces as a prioritisation problem. Roadmap decisions that made sense when activation was the primary signal may not make sense when expansion revenue is where the model actually needs to perform. Understanding which user you are optimising for, and whether that is the user who drives the business outcome, is not a strategic question someone else will answer for you.

The PLG playbook was never meant to be permanent. The teams doing well in 2026 are the ones who noticed that first.