From Freemium to Flywheel: How Product-Led Growth Reinvented Itself

In 2011, a product manager at a mid-sized software company stood in front of her board and presented something unusual: a growth plan with no sales team. The product would sign up users on its own. A free tier would do the convincing. Word of mouth would handle distribution. The board called it naïve. Three years later, the company had a hundred thousand users and a buyout offer.

That product manager wasn't wrong. She was just early. What she was describing had no name yet, no benchmark reports, no VC thesis around it. It was simply a bet that a good product, made easy to try, could grow faster than a quota-carrying sales force.

That bet is now worth hundreds of billions of dollars. And it has changed — profoundly — in ways that most people discussing it in 2025 still haven't fully processed.

Here is the full arc.

The roots of what we now call Product-Led Growth trace back to the early 2000s, when the first wave of SaaS companies was building subscription software and wrestling with a fundamental problem: enterprise software had always been sold by people, not experienced by them. Procurement cycles ran for months. Demos were theatrical. Lawyers negotiated contracts.

Atlassian, founded in 2002, took a different path almost by accident. Building developer tools — Jira for issue tracking, Confluence for documentation — the company targeted individual engineers who could install the software themselves, try it without a handshake, and pull it into their organizations through actual use. For years, Atlassian had essentially no field sales force. Revenue grew anyway, because the product did the convincing work that salespeople couldn't.

Dropbox arrived in 2008 with a similarly subversive premise: give users free cloud storage, make sharing a file inherently an act of product distribution, and let the behavior spread. Their referral program — extra storage in exchange for each friend who joined — produced a 3,900% increase in users in fifteen months. They were not spending on ads. They were engineering virality directly into the product experience.

These companies were operating on instinct rather than doctrine. The playbook didn't exist yet. They were writing it.

The term "Product-Led Growth" was formalized in 2016 by Blake Bartlett, a partner at venture capital firm OpenView. The naming mattered because it gave a framework to something companies had been doing intuitively — and it signaled that VCs were paying attention. PLG was no longer just a product philosophy. It was a go-to-market thesis worth building a fund around.

The timing was not coincidental. By 2016, Slack had demonstrated something remarkable: a team messaging tool could spread through entire organizations without a single outbound call. In its first year, Slack reached over 285,000 daily active users through pure word-of-mouth and product virality, without hiring an outbound salesperson until three years after launch. Zoom was following a similar path — a zero-friction video call that required no account to receive, which meant every meeting was also a product demo.

The market capitalization of PLG companies grew from $21 billion in 2016 to $687 billion by 2020. That figure alone explains why the term entered the vocabulary of every SaaS founder, investor, and product manager in the late 2010s.

New metrics emerged alongside the strategy: Product-Qualified Leads, or PQLs — users who had reached a defined activation threshold and were therefore more likely to convert than any form of marketing-generated lead. Studies found that PQLs converted at five to six times the rate of traditional marketing-qualified leads. The product was not just growing the company. It was producing better-quality buyers than the marketing department.

PLG 1.0 — the model that Dropbox, Slack, Zoom, and Calendly embodied — was built around a single organizing principle: eliminate every barrier between the user and the first moment of value.

The mechanics varied by company but followed a recognizable structure. Offer a free tier or free trial that requires no credit card and no sales conversation. Design onboarding so a user could reach their "aha moment" — the specific point at which the product's value became undeniable — within the first session. Build sharing or collaboration natively into the product so that every user who finds value automatically introduces new users to the experience.

Calendly crystallized this pattern almost elegantly. Every scheduling link a user sent was an advertisement. The recipient who clicked the link experienced the product without signing up — and many of them signed up anyway because the experience spoke for itself. As OpenView observed, every Calendly invite started a viral loop. The company grew to millions of users without a meaningful outbound motion.

The weakness of PLG 1.0 was that it was, fundamentally, a user acquisition strategy. It was extraordinarily efficient at generating signups and even at converting individual users to paid plans. Where it struggled was in moving upmarket — in capturing the large enterprise deals with six-figure contract values that required navigating procurement committees, security reviews, and executive sign-off. A freemium product that individual developers loved was not necessarily the product that a Chief Information Officer would approve for company-wide deployment.

Many PLG companies hit this ceiling in the early 2020s. They had millions of users and respectable revenue from self-serve conversions, but they were leaving the enterprise segment largely to competitors who were willing to run a traditional sales motion. The PLG trap — a term coined by Oliver Jay, a former CRO at Asana and Dropbox — described companies that over-relied on self-serve growth and then found themselves unable to scale into the deals that would define the next phase of the business.

The maturation of PLG brought with it a more sophisticated analytical framework. The first generation had often measured success through vanity metrics — signups, monthly active users, top-of-funnel traffic. The second generation of PLG operators understood that these numbers were misleading without a clear line of sight to activation, retention, and expansion.

Activation — the rate at which new users reached the point of real value — became the critical early signal. The benchmark for average SaaS companies settled around 33%. The top ten percent of PLG companies achieved activation rates above 65%, which meant that two out of three new signups reached the moment at which the product had made its case.

Net Revenue Retention, or NRR — the measure of revenue growth or contraction within an existing customer base — became the single most important metric for evaluating the health of a PLG business. A company with 120% NRR was growing purely from its existing customer base without adding a single new logo. Slack grew this way: teams that started with five users became teams of fifty, then five hundred. Each expansion was revenue the company didn't have to acquire through marketing or sales.

The industry also reckoned with the limits of "unlimited free." By 2025, the dominant model had shifted from generous open-ended freemium to strategic freemium — free tiers designed to deliver genuine value while creating natural friction points that made upgrading feel inevitable rather than forced. Calendly introduced caps on meeting types. Notion limited collaborative features on the personal plan. The calculus had changed: free users are not cheap in a world where compute costs are real, and the conversion path from free to paid must be engineered deliberately.

The most instructive strategic evolution of the mid-2020s was not any company's decision to adopt PLG. It was the wave of mature PLG companies that quietly built sales teams on top of their self-serve foundations.

Atlassian added an enterprise sales layer after years of pure PLG. Slack built an enterprise division that pursued large-account contracts. Notion hired its first marketing and sales team well into its growth story. Zoom added a sales-assisted motion for its larger accounts. None of these moves represented a retreat from PLG. They represented its maturation — the recognition that the most efficient path to large enterprise revenue ran through the product first, and then through a human who could navigate the procurement reality on the other side.

The hybrid model — which the industry began calling Product-Led Sales, or PLS — became the dominant architecture for B2B SaaS companies above a certain scale. The product drives adoption from the bottom up. Individual users become internal champions. When an account shows sufficient product engagement, a sales representative steps in to close the enterprise deal that the product has already half-sold. By 2026, companies running hybrid motions reported twice the profitability of those running either pure PLG or pure sales-led strategies.

Then the rules changed again.

PLG 2.0 is not simply PLG with an AI feature bolted on. It is a structural transformation in what the product does for the user. PLG 1.0 products removed friction so users could build things themselves. PLG 2.0 products do the building — generating the first draft, completing the first task, producing the first output — so that users arrive at value before they have had a chance to disengage.

Cursor reached $100 million ARR in twelve months, then $2 billion ARR by February 2026, with zero marketing spend. Lovable, a Swedish AI app builder, crossed $100 million ARR in eight months with forty-five employees — generating over $2.2 million in revenue per headcount. These are not just fast-growing companies. They are evidence of a different category of product-market fit, one in which the user's first interaction with the product is already an experience of delivered value rather than a promise of future value.

The implications for onboarding, pricing, and retention are significant. When the product does the work rather than enabling the user to do the work, the time-to-value collapses from hours or days to minutes. The activation rate problem — the long-standing challenge of getting users to the aha moment before they churn — shrinks considerably when the aha moment is the first thing they encounter.

Pricing evolved in parallel. Per-seat models, which charged for access regardless of usage, began giving way to usage-based and task-based structures. Work as a Service — charging for tasks completed rather than seats occupied — aligned price directly to the value delivered. The most ambitious operators were beginning to experiment with Results as a Service: charging for outcomes rather than activities, which aligned incentives completely but also required clear, attributable measurement of those outcomes.

The next evolution is already underway. PLG 3.0 describes products where the primary user is not a human but an AI agent — systems that receive a goal, execute a workflow, and return a result for human approval. The human's role shifts from builder to reviewer.

For product-led growth, this creates a fundamental question: if users don't "use" the product in the traditional sense, what does activation mean? What does virality look like when there is no sharing action to engineer? What does retention mean when the relationship between user and product is mediated by automation?

These questions don't yet have settled answers. But the companies working through them — a small set of developer infrastructure tools and AI-native platforms — are sketching the next chapter of a strategy that has already rewritten how software companies are built.

Three things have remained constant across all three generations of PLG, and they are worth restating because they are easy to lose in the noise around AI, pricing models, and growth mechanics.

First: the product must deliver undeniable value in the first session. Not promised value. Not the value described in the marketing copy. Actual value, experienced directly, before the user has any reason to trust the company.

Second: the motion must remove friction from the path to that value. Every step added between signup and the first moment of success is a tax on growth. The most successful PLG companies treat friction not as a minor inconvenience but as an existential threat.

Third: user success is the growth engine. The company that grows because its users are achieving their goals has built something durable. The company that grows through promotional mechanics and aggressive conversion pressure has built something fragile.

Everything else — the AI features, the pricing models, the hybrid sales layers, the agentic workflows — is in service of those three constants. PLG is not a tactic. It is a commitment to making the product the best argument for itself.