With the rise of sophisticated AI coding assistants, a narrative has emerged suggesting the imminent demise of the software-as-a-service industry. Investors are spooked, and tech valuations for giants like Atlassian and Figma have taken a hit. Yet, amidst this uncertainty, our guest, Vijay Raina, a leading expert in enterprise SaaS technology and software architecture, argues that these fears are wildly exaggerated. He contends that specialized SaaS businesses possess a powerful, defensible moat built on deep industry knowledge and compounding user feedback. Today, we’ll explore why the DIY AI approach has significant hidden costs, how SaaS companies can stay “five steps ahead” of in-house solutions, and how they can leverage AI to widen that lead.
You argue that a specialized SaaS business will always be “five steps ahead” of a custom AI-built tool. How does the compounding feedback from hundreds of global merchants create this advantage, and can you share a specific example of a feature developed from this insight?
It’s a powerful dynamic that a single in-house team simply can’t replicate. Imagine having the constant, real-time feedback of literally hundreds of merchants from over 50 countries, all telling you what they need to win more sales. That isn’t just data; it’s a living, breathing roadmap. Every piece of guidance, every feature request, every pain point from our user base, which now exceeds 800 clients, is information that we take and compound into the product. This creates a cycle of improvement that accelerates over time. An in-house team is building in a vacuum, solving for their own problems, while we’re solving for the entire industry’s problems before many of our clients even encounter them.
Some estimate that building a refined tool with AI could cost hundreds of thousands in compute tokens. Can you elaborate on this hidden cost versus a subscription fee and detail the specific, niche functionalities that an in-house team would struggle to replicate without deep industry knowledge?
Absolutely. The allure of “vibe coding” a custom solution with AI is strong, but the reality is a significant financial drain that goes far beyond a predictable subscription fee. We’re talking about a potential cost of hundreds of thousands of dollars just in AI compute tokens to even approach the level of refinement and functionality we offer. But the cost isn’t just monetary. An in-house team, no matter how skilled, lacks the sector-specific experience. For example, in the neon sign industry, we’ve built quoting tools that account for regional material costs, international shipping complexities, and design mock-ups that reflect very specific manufacturing constraints. An internal team wouldn’t even know to ask those questions until months of costly trial and error. We’ve already built those solutions because our global merchants told us they needed them.
Prominent software companies have seen valuations fall amid fears that clients will use AI to build their own tools. How do you articulate your company’s defensible moat to investors, and what key metrics do you highlight to prove the long-term value of your specialized platform?
It’s a conversation we’re having constantly, and the key is to shift the focus from the technology to the network. Our defensible moat isn’t just our code; it’s the ecosystem we’ve built. I point investors to the compounding value of our feedback loop from over 800 clients worldwide. I highlight the tangible results, like facilitating nearly $17 million in merchant sales through our platform in a single year. That’s not just a software tool; that’s a growth engine for an entire industry. I articulate that while anyone can try to build a tool, no one can replicate our user base’s collective intelligence overnight. As Mike Cannon-Brookes from Atlassian said, it’s simply more efficient for businesses to buy pre-canned, expert solutions.
You mentioned that an in-house team could eventually build a competing product with AI. How do you use AI tools within your own business to accelerate development and widen the gap, ensuring that by the time a competitor launches, your product has already evolved significantly?
This is the crucial point that often gets missed in the debate. We’re not fighting against AI; we are avid users of it. We see AI tools as offering infinite possibilities to accelerate our own roadmap. By integrating AI copilots and other development tools internally, we’re able to process that vast amount of user feedback and deploy new features and improvements faster than ever before. So, while an ambitious internal team is spending a year and a fortune in compute tokens to build their version 1.0, we’ve already gone through several major iterations. By the time they build it and launch, we are already five steps ahead, solving the next set of problems our industry will face.
What is your forecast for the SaaS industry over the next five years as AI-powered coding assistants become even more sophisticated and accessible?
I believe we’re going to see a great divergence. On one hand, generic, horizontal SaaS tools may face pressure as basic functionalities become easier to replicate in-house. However, for specialized, vertical SaaS businesses, this is an incredible opportunity. The winners won’t be those who simply build custom tools with AI, but those who leverage AI to deepen their expertise and serve their niche communities more effectively. The human ego has driven large companies to build their own software for decades, and as Niki Scevak of Blackbird noted, that has almost always been a bad idea. AI doesn’t change that fundamental business logic. The future of SaaS is more specialized, more integrated, and more valuable than ever, precisely because it will be powered by, not replaced by, artificial intelligence.
