Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

Rob Walling cites Peter Levels paying for ScrapingBee because AI cannot manage proxies and CAPTCHAs. Anar Volset compares email deliverability: one customer cannot hire 25 people to fight spam and IP issues. The cost only makes sense amortized across many customers, so shared infrastructure remains sticky.

Related Insights

While AI lowers the bar for building simple internal tools, it doesn't solve for the complexity of mission-critical systems. The need to manage intricate workflows across multiple departments (IT, sales, finance, legal), handle integrations, and ensure compliance makes specialized vendors a safer choice.

A primary barrier to deploying autonomous AI agents isn't their intelligence, but the internet's existing infrastructure. Current systems, with rate limits and spam filters, are not designed for high-frequency agentic activity and often block them, limiting their ability to operate effectively.

While AI can replicate the functionality of a SaaS tool, it doesn't replicate the company infrastructure: sales, customer support, trust, and brand. With venture funding for new SaaS startups drying up, it's harder than ever for new entrants to reach critical mass, thus protecting established incumbents.

While AI can build an initial version of a software product instantly, the true, defensible value lies in the ongoing maintenance, support, and reliability. Customers will always pay for a product that is actively maintained and improved over time.

Frontier AI companies still pay for legacy SaaS not because they can't build alternatives, but because it's a poor allocation of scarce, expensive engineering resources to replace established, functional tools.

CEOs of platforms like ZocDoc and TaskRabbit are not worried about AI agent disruption. They believe the immense complexity of managing their real-world networks—like integrating with chaotic healthcare systems or vetting thousands of workers—is a defensible moat that pure software agents cannot easily replicate, giving them leverage over AI companies.

Many developers believe tweaking prompts and logic ('harness engineering') is the hardest part of building agents. The real bottleneck, however, is scaling, reliability, and managing production infrastructure—a common miscalculation that managed services aim to solve.

For years, businesses have focused on protecting their sites from malicious bots. This same architecture now blocks beneficial AI agents acting on behalf of consumers. Companies must rethink their technical infrastructure to differentiate and welcome these new 'good bots' for agentic commerce.

With AI commoditizing code creation, the sustainable value for software companies shifts. Customers pay for reliability, support, compliance, and security patches—the 'never ending maintenance commitment'—which becomes the key differentiator when anyone can build an initial app quickly.

The ability to rapidly build custom software with AI is tempting. However, the ongoing maintenance and data quality assurance are the core business of SaaS companies. Buying a dedicated tool like a CRM often provides more value and less overhead than a custom-built solution, even with AI assistance.

Hard-to-run infrastructure beats AI code generation | RiffOn