Why HubSpot’s Co-Founder Is Betting Against the SaaSpocalypse

Why the software opportunity is growing, not shrinking, and how to stay off the Yellow Brick Road the frontier labs are racing down.

With Dharmesh Shah

"The actual thing that's valuable is the work that's happening. The interface is a means to an end." — Dharmesh Shah

Dharmesh Shah, on the episode

Dharmesh Shah, co-founder and CTO of HubSpot, sits down with a16z partner Joe Schmidt to work through a question that's been circulating since Schmidt's essay on the Yellow Brick Road, the idea that frontier AI labs are racing down one well-lit path while leaving enormous territory untouched on either side. The pop version of this moment says software is getting commoditized into oblivion by chat interfaces and ever-smarter models. Shah's actual operating system says the opposite: software was only ever built to solve the class of problems it could solve cheaply, and AI just expanded that class dramatically, which means there's more room to build, not less. He draws this out through HubSpot's own 20 year history, from a narrow law-firm pitch that got abandoned in weeks to a CRM built because 60 percent of paying customers had no CRM at all. What follows is a working model for picking ideas, defending a market against labs with more capital and compute, and deciding what to build when the interface itself stops being the product.