Why Canopy’s Self-Serve PLG Experiment Attracted Stanford Kids Instead of Enterprise Buyers

Why Canopy’s Self-Serve PLG Experiment Attracted Stanford Kids Instead of Enterprise Buyers

The PLG playbook looks bulletproof on paper: remove friction, let users self-serve, charge based on usage, scale through virality. It worked for Slack. It worked for Notion. For Canopy, it spectacularly failed. In a recent episode of Category Visionaries, Matt Bivons, CEO of Canopy, dissected why product-led growth doesn’t work for every B2B product—and why trying to force it cost them valuable time chasing the wrong customers.

The Seductive Logic of Self-Service

“The other piece that I spectacularly failed at was trying to do a self serve lending as a service,” Matt admits without hesitation. The vision made sense: build a freemium model, let companies sign up and experiment, charge per API call like Twilio. Lower customer acquisition costs, faster sales cycles, more predictable revenue growth.

The reality delivered a harsh lesson in market dynamics. “It attracted the wrong type of customer. So were getting kids from Stanford in their dorms trying us out. And lending is really hard. Like you have to know what you were doing and that was not our customer,” Matt explains.

This wasn’t a marketing problem or a positioning issue. It was a fundamental misalignment between go-to-market strategy and product complexity. When your product requires deep domain expertise just to evaluate properly, self-service becomes a filter that selects for exactly the wrong buyers.

The Pricing Model That Confused Enterprise Buyers

Beyond attracting unqualified users, Canopy discovered that usage-based pricing—the cornerstone of PLG economics—collapses under lending’s inherent complexity.

“Additionally to that charging per API call, similar to how Twilio does, is very confusing to most people when you’re dealing with lending because the amount of contacts that you get matters based on the lending product you have,” Matt notes.

The problem wasn’t just confusion—it was fundamental unpredictability. Matt breaks down why: “For example, if you have a student loan, you’re never really calling in to your servicer. It just you set up autopay and it deducts every month. Same thing with mortgages. But if you have a revolving card or a debit card and you get declined the point of sale, you’re absolutely jumping on the phone.”

Contact rates vary by 10x or more depending on product type. A company evaluating Canopy’s APIs can’t model their costs because usage patterns depend on variables they won’t know until months after implementation. When your ICP can’t budget for your product, pricing becomes a sales barrier rather than a sales enabler.

The Multi-Stakeholder Reality Nobody Talks About

The deeper issue with PLG for complex fintech infrastructure isn’t about features or pricing—it’s about how enterprise buying decisions actually work.

“What we found out was that we needed to obviously talk to various stakeholders at these companies, which was all outbound,” Matt explains. Lending infrastructure decisions don’t get made by individual contributors signing up for free trials. They involve compliance teams ensuring regulatory requirements are met, engineering leaders evaluating technical architecture, finance executives modeling cost implications, and product managers assessing feature completeness.

No single person can sign up for Canopy’s freemium tier and make a buying decision. The sales cycle isn’t something to optimize away with better onboarding—it’s inherent to how these purchases work.

When PLG Actually Makes Sense (And When It Doesn’t)

Matt’s experience led him to a nuanced view that challenges the PLG evangelism prevalent in B2B circles. “I don’t think that product led growth is right for every company, every product,” he notes.

The critical insight: your go-to-market motion must match your market’s buying behavior, not your aspirational growth model. “You have to think about your distribution and your market and customer in mind. So you can’t disconnect what you’re building from who you’re selling it to and how you sell it and how you go to market will change based on the industry you’re in and the customers that you sell to,” Matt explains.

For fintech specifically, structural factors limit PLG viability. “I think that product led growth is a challenge. Right? Certainly there are products that are seat based… That can definitely happen. But there isn’t necessarily the virality or referral engine that happens from traditional SaaS PLG when you’re talking about fintech,” Matt says.

Lending infrastructure isn’t viral. Compliance officers don’t invite colleagues to try experimental APIs. Finance teams don’t share freemium tools on Twitter. The social dynamics that power PLG growth loops simply don’t exist in regulated financial services.

The Evolution to High-Touch Enterprise Sales

After the PLG experiment failed, Canopy rebuilt their go-to-market from scratch. “It evolved from just me into now we have chief revenue officer and full sales and marketing team who do this,” Matt notes.

But it wasn’t just about adding salespeople. The real shift was understanding the buyer journey’s inherent complexity. “It also mattered around how many touches we actually needed to move people through our funnel,” Matt explains.

Canopy discovered two distinct buyer segments. Some have urgent pain—reconciliation nightmares, systems that can’t support new products. These brownfield deals move relatively fast. But they’re the minority.

“The majority of companies, it’s a year long journey and we need to have high trust and build relationships with them across many months,” Matt reveals. Year-long sales cycles require a completely different approach than transactional PLG. “That comes in the form of not being, you know, overselling, but really listening and understanding the company, their business model, their future,” Matt says.

Content Marketing as the PLG Replacement

With self-service dead, Canopy needed a way to stay engaged during year-long enterprise sales cycles. Their answer wasn’t more aggressive outbound—it was content.

“It also comes from creating a lot of content to help them understand when is the right time to use us,” Matt explains. This content serves a fundamentally different purpose than PLG’s in-product education. It’s not about showing users how to get value—it’s about helping prospects understand when their current pain justifies switching.

“I would say the combination of content marketing and direct sales, which is all about building trust, is where we are right now. And that is what is most high converting and working for us,” Matt says. Content marketing and enterprise sales aren’t competing strategies—they’re complementary motions that work together for complex B2B products.

The Framework for Choosing Your GTM Motion

Matt’s failed PLG experiment offers a practical framework for founders choosing between self-service and high-touch sales:

Does your product require domain expertise to evaluate? If yes, self-service will attract unqualified users who waste support resources without converting.

Do buying decisions involve multiple stakeholders across different departments? If yes, individual sign-ups won’t lead to enterprise deals.

Does usage vary dramatically based on use case? If yes, usage-based pricing creates unpredictable costs that enterprise buyers can’t budget for.

Is implementation and customization more important than instant time-to-value? If yes, the friction you’re trying to remove with PLG might actually be necessary.

For Canopy, every answer pointed away from PLG and toward high-touch enterprise sales.

The Cost of Wrong-Channel Strategy

Failed go-to-market experiments cost more than just wasted time and money—they cost opportunity. While Canopy was optimizing for self-service sign-ups, their ideal customers weren’t finding them through that channel.

“Obviously fail a lot, fail fast, figure it out, run experiments and do what you feel is working and then double down on that,” Matt advises. The corollary: be willing to kill experiments fast, even when they’re experiments you really wanted to succeed.

The self-serve dream is seductive—lower CAC, predictable growth, venture-scale returns. But for complex B2B infrastructure products, especially in regulated industries, the high-touch reality might be your actual competitive advantage, not a weakness to engineer away. Sometimes the thing that doesn’t scale is exactly what should scale.