Takeaways from Vertex Spring 2026, a semi-annual gathering of Vertical SaaS operators.
Spend enough time reading the headlines, and you might believe SaaS is in trouble. AI can supposedly replicate any product in a weekend. Investors are getting cold feet over subscription revenue. The dorm-room college kid with a Cursor subscription is coming for your customers.
But spend a day at Vertex, a conference for vertical software founders, operators, and investors, and you’ll hear something different. You’ll hear from people who’ve actually built category-defining businesses in specific industries, and who have looked past the headlines about what the AI era actually means for their companies.
The theme this year: durability. The SaaSpocalypse is certainly real for large horizontal enterprise SaaS products. But for vertical SaaS operators deeply understand their industries, it’s an opportunity.
Here’s what the operators in the room had to say.
The Moat Was Never the Software
The most repeated line of the day came from Matt Brown on the Built For This: How vSaaS Wins the AI Era panel: “It was never about the software […] It is about the vertical and the trust you build.”
Alex Jekowsky, co-founder and CEO of Cents, a vertical SaaS platform purpose-built for the laundry industry, shared how he built trust. Cents has hardware sitting directly on laundromat machines. That hardware captures operational data no one else has, integrated so tightly into the physical workflows of its customers that replication is nearly impossible. A 20-year-old with a great LLM is not going to go install hardware in hundreds of laundromats, build trust and deep relationships with operators over years, and develop the domain expertise to know what that data means.
Which is the real question the panel kept returning to for early-stage vertical SaaS founders: what is that 20-year-old not going to do? Build your moat there.
For Cents, it’s on-premise laundromat hardware. For Clio, serving the legal industry, it’s state-specific court filings software, which needs to be customized for each jurisdiction (and needs to be accurate).
As a vertical SaaS operator, thinking deeply about your moat is a useful forcing function that directs attention away from software features, which developers can replicate easily, and toward the harder, stickier things: regulatory relationships, hardware integrations, proprietary operational data, and the decade of domain knowledge that tells you what to do with all of it. While this strategy diverges from collective wisdom a decade ago, when a vertical SaaS platform would’ve been discouraged from building hardware, that’s now a key differentiator.
Joshua Silver, founder and CEO of Rainforest, framed it this way in his opening keynote: “Vertical software doesn’t just sit on top of workflows. It is the workflows. And that’s the last mile that AI will have a really hard time displacing.”
The TAM for Vertical SaaS Is Actually Expanding
What might be bad for large horizontal enterprise platforms could serve vertical SaaS platforms well. Startups can build software for less cost (and faster) than ever before, so vertical SaaS platforms can now profitably reach customer segments that were previously uneconomical to serve.
AI lowers the cost to build, yes. But it also lowers the cost to serve, which means the addressable market for vertical software is growing, not shrinking. Categories that previously required enormous engineering investment to serve (not to mention sell to or support) can now be entered at a fraction of the cost, opening up new verticals and customer segments.
This means you can build more specialized vertical SaaS software, like those specifically designed for lawn care or asphalt pavers, to serve business owners better than ever before. But you must couple this code with boots on the ground: as the CEO of Cents said, that means meeting laundromat owners onsite, or in the case of Cubby, building relationships self-storage owners and operators.
“Regardless of your vertical,” Silver said, “you can still build a thriving payments business.”
The optimized companies in the benchmark (available soon) are generating 3.5x the payments revenue of their early-stage peers. The message for operators who haven’t yet invested seriously in embedded fintech: the window is open, but the companies that start now will compound their advantages over those who wait.
AI Is a Feature, Not a Product
Panel after panel at Vertex reinforced the same sequencing: vertical pain point first, technology second.
Susanne Sandler, SVP & GM of Fintech at Mews, a vertical SaaS platform for hotels, articulated the principle with a specific example during the Embedded Fintech: What We Learned the Hard Way panel: for economy hotels, technology with AI built-in like check-in kiosks reduces labor costs; for luxury hotels, that same technology frees up staff to spend more time with guests. The use case is different in each segment. The technology is a means to an outcome that was already meaningful to the customer, not a product in itself.
The Jane App story is perhaps the cleanest illustration of this principle playing out at scale. Ali Heckler and Trevor Johnston built a practice management platform for healthcare clinics that now approaches $200M in ARR and a $1.8B valuation, without ever taking primary capital. Payments weren’t a strategic initiative. They were just what the product needed.
“Jane payments started because it made sense for Jane to handle payments,” Trevor explained. “It wasn’t a business decision to go try to have a fintech thing. It was just natural. We’re doing everything else to do with the checkout. Wouldn’t it also make sense for Jane to actually process the payment?”
That unplanned move into payments now represents 36% of revenue, and growing. The pattern holds: companies that embed fintech because it serves the customer better build far more defensible products than companies that bolt on payments as a revenue play.
Trust Compounds
Across every panel, “trust” kept surfacing as the real variable that separates the durable companies from the ones that won’t survive the AI era.
Matt Engfer, co-founder and CEO of Cubby, a self-storage platform, framed AI as an opportunity to lean into the customer relationship rather than displace it. The self-storage operators who will win aren’t just the ones with the best product; they’re the ones who become the trusted resource their customers turn to when they’re trying to figure out how to use AI in their own businesses. Own that relationship, and you’re not just sticky; you’re irreplaceable.
The Vertical SaaS founders on the afternoon panel extended this argument in a different direction: trust isn’t just a customer relationship strategy, it’s the core design challenge for AI-native products.
Mitchell Troyanovsky, co-founder of Basis, an AI platform that deploys agents at accounting firms, has spent a lot of time thinking about how customers build trust with AI that isn’t deterministic. “Humans are really used to working with nondeterministic systems,” he said. “It’s just those systems are normally their co-workers, not their computers. If a human you hired made a mistake one out of 20 times, it’s pretty good, right?”
The implication: the problem isn’t that AI makes mistakes, but that products are designed around a promise of deterministic accuracy that was never realistic. As he says, “we’re dealing with intelligence, not software,” and need to set expectations appropriately. Build the product experience so customers can incrementally extend trust to agents, the same way they’d onboard a new hire, and the adoption follows.
Ian Weng, Chief Strategy Officer at Elise AI, a vertical AI company automating property management workflows, added the physical-world dimension: “We intersect with the physical world. I think that’s probably one of the final frontiers in terms of the larger labs. Even if a larger lab can do […] 50% of what we do, customers at the end of the day just want their problem solved. They want ideally one provider to solve 100% of their problems.”
Foundation models can’t manage maintenance tickets, door access codes, and on-site tours. Like hardware, the companies that go deep enough into the physical operational layer of their verticals are building moats that no general-purpose AI can easily cross.
The Next Horizon: Agentic Commerce
If the first half of the day was about durability, the second half was about offense.
Ryan Pritchard, SVP of Engineering and Agentic Commerce at PayPal, walked through three stages of how agentic commerce is unfolding: computer-using agents that interact with websites the way a human would; API-layer access that lets AI orchestrate workflows beneath the human interface (where most activity is right now); and fully autonomous commerce, where the human sets intent and steps entirely out of the loop.
The numbers are striking: $3 billion in AI-driven commerce happened in just the five days of Black Friday and Cyber Monday — an 11x increase year-over-year. Analysts project that 25% of all e-commerce will be conducted agentically within three and a half years.
For vertical SaaS platforms, this isn’t just a trend to monitor. It’s an opportunity to compound the advantages they’re already building.
“I actually think vertical SaaS has even a bigger opportunity for disruption and to win against your competitors,” Pritchard said. “If you think about the vertical, there’s more parts of the workflow in service-based businesses that happen on your platforms. If your systems that are interacting with an LLM can streamline that really, really well and tell a great story about what types of services you offer, you have a huge opportunity there.”
The multi-step, context-rich processes that define vertical SaaS, from scheduling to compliance, payments, and field operations, combine all the necessary ingredients for AI agents to create real value for customers. The platforms that are already embedded in those workflows don’t need to reinvent themselves for the agentic era. They just need to make those workflows machine-readable, stand up an MCP server, and let their decade of proprietary data do the work.
Vertical Software Moats are Primed for AI
Vertical SaaS companies building trust with their customers don’t need to worry about vibe-coding competitors. They’re going deeper into the physical operations of their customers, layering financial products on top of relationships that took years to earn, positioning themselves as the trusted guide for what comes next.
As Joshua Silver put it in his opening, “All of you have built highly respected brands and trust in your specific verticals. That doesn’t just allow you to sustain and survive in an AI era. You’re the one who gets to define what it looks like.”
The structural advantages of vertical software are real, particularly paired with a moat like hardware. The so-called “SaaSpocalypse” will clear out the undifferentiated players and leave more room for the companies that actually know their industries and their customers.
Watch the Sessions
- Opening Keynote — Joshua Silver, CEO, Rainforest
- Jane App: The $1.8B Bootstrapped Playbook — Ali Heckler & Trevor Johnston
- Built For This: How vSaaS Wins the AI Era — Matt Brown/Matrix, Gavin Nachbar/Aiwyn, Matt Engfer/Cubby, Alex Jekowsky/Cents
- Embedded Fintech: What We Learned the Hard Way — A.J. Axelrod/Clio, Michel Rbeiz/Toast, Susanne Sandler/Mews
- Vertical AI: Building the New System of Work — Mitchell Troyanovsky/Basis, Ian Weng/Elise AI, Max Brenner/WithCoverage (moderated by Nihar, Better Tomorrow Ventures)
- Agentic Commerce: Preparing Your Platform — Ryan Pritchard, SVP Engineering & Agentic Commerce, PayPal
