Maria Thomas on Product-Led Growth: Why Buffer Prioritizes Product Experience

The $25 Price Point That Almost Killed the Company

Maria Thomas and her co-founders made what she calls "a common startup mistake" — they built the product the way they thought it would be used, not the way their actual customers needed to use it. As builders who had only ever bought software on credit cards for a couple hundred dollars a month, they naturally assumed everyone else operated the same way. They launched with a $25 monthly price point, oriented the entire experience around frictionless self-serve, and waited for the growth to kick in.

It didn't. The company started running out of money.

We were running out of money, so it's like something's not going great. We read about what it feels like to have product market fit and it's like, oh, things are on fire, people want more of your product, everything's going well, up to the right. And we're like, ah, that's not how it seems to be for us.

— Maria Thomas

The problem wasn't the product itself. The problem was that Thomas and her team had designed a go-to-market motion for hobbyists and first-time founders when their actual value proposition served sophisticated product teams at companies with 100 to 1,000 employees. These weren't people looking to swipe a credit card. These were buyers with evaluation processes, POC requirements, and procurement departments. The mismatch between product positioning and customer reality was existential.

When Advisors Tell You to Kill Self-Serve Entirely

Experienced advisors gave Thomas stark advice: completely pivot to sales-assisted. Put a "Request Demo" button on the website. Remove self-serve signup entirely. For someone from a product-led background, the suggestion felt radical, almost heretical.

Thomas didn't follow the advice to the letter, but she did start offering demo options. More importantly, she began mapping the buying journey her actual customers were on. They wanted to get started quickly themselves, but then they'd get stuck. They needed help with specific technical questions. They needed their data flowing correctly and integrations verified. The solution wasn't to abandon self-serve completely — it was to build a hybrid process that met customers where they actually were.

We didn't quite go down that route, but we did at least offer that and say, okay, hey, if you do want to book a demo, or if you do want to talk to somebody, we can do it. And then we tried to think about what does that process itself look like, you know, engineered it in a way, like what's going to be convenient?

— Maria Thomas

The shift required Thomas to learn an entirely new skillset. She hired a sales coach and went through role-playing exercises that revealed how much she didn't know. Multi-threading deals, avoiding procurement surprises, preventing sticker shock — none of it was intuitive for someone with an engineering and product background. But the discomfort was necessary. Once the new motion started working, the company moved from tens of thousands of dollars in annual contracts paid upfront to sustainable revenue growth.

Product-Market Fit Is Never Binary

Thomas rejects the mythology around product-market fit as some moment of explosive, undeniable demand. She's never experienced it that way, and she doesn't think most founders do either. For her, the signal came from win rate. When the company started closing deals against competitors more often than not, when customers started upgrading and paying more over time, that's when she knew something fundamental had shifted.

It's always with product market fit, it's never binary. It's always like gradations of it. So I think there's more to come. But yeah, it feels like it's working now. And we have different problems, which is like, how do we scale faster?

— Maria Thomas

The problems changed from "are we going to close anything?" to "how many are we going to close?" That transition — from existential uncertainty to execution challenges — marked the real inflection point. Thomas also points to a Stripe report showing that PLG really kicks in after $10 million ARR for SaaS companies, a data point that validated her experience. The early-stage worship of pure self-serve often ignores the reality that most companies need a hybrid approach until they reach meaningful scale.

The Dangerous Framing of Sales-Led vs. Product-Led

Thomas pushes back on the binary framing of sales-led versus product-led growth. The either/or mentality misses the point. The real question is: where can you incrementally remove handholding and make experiences more self-service? Maybe it's in the evaluation phase. Maybe it's a sandbox environment. Maybe it's an upsell flow. The goal isn't to pick a camp but to let customers buy the way they want to buy.

She's also clear that building the product is harder than training a go-to-market motion, though she acknowledges her builder bias. Moving from an enterprise product to a PLG motion is especially difficult because the culture gets built around human support. Those silos calcify. Going the other direction — from PLG to adding sales-assist — has its own challenges, but it's more about learning than cultural transformation.

I don't want to come across against PLG. Our product helps other companies enable PLG. Last thing you want to do is send a message that PLG doesn't work. So it does work. And that itself is also a gradation.

— Maria Thomas

Thomas believes early-stage companies should operate more like concierge services, individually handholding customers to learn deeply about their needs. That learning informs what can eventually be productized and scaled. The S-curve matters: high-touch in the beginning, self-serve at scale, but always oriented around the customer's actual journey, not the founder's assumptions about how buying should work.

Coaching Customers to Success, Not Just Selling Software

One of the hardest lessons Thomas learned was that sophisticated products require customer education. When customers first started using the product, they built ugly, unoptimized experiences — the equivalent of version one marketing emails that everyone immediately deletes. The technology was new, and users didn't yet know how to wield it effectively.

The temptation for a product-led company is to let the product speak for itself. But Thomas realized that coaching customers led to drastically better outcomes. When the team invested in guidance, customers showed real numbers, stuck around longer, and became champions. The challenge was that coaching doesn't scale the same way pure self-serve does. It required a different engagement model, one that prioritized long-term success over short-term acquisition velocity.

Part of it was that we needed to coach people. And when we were able to coach them, they were way more successful. But if you have to coach somebody, that's a different model of engagement than just like, here's the website, go read about it, go buy it.

— Maria Thomas

This insight reshaped how Thomas thought about go-to-market. The company narrowed in on their ideal customer profile, defined personas more clearly, and mapped the buying journey in detail. The result wasn't just a better sales process — it was a model that actually met customer needs. Thomas doesn't frame it as pushing sales. She frames it as building the experience that helps customers succeed. When that alignment happens, everyone wins.