via jokof
Wall Street is dead wrong about the death of SaaS. The narrative that AI will let every company build their own software for pennies, misses how enterprise business actually functions.
SMB revenue is a rounding error.
The bear case focuses on small businesses using AI to replace simple tools, but major SaaS players live on the Fortune 500. Large enterprises don’t buy software just for features. They buy SOC2 compliance, HIPAA, and legal accountability. They need one throat to choke when things break. An AI generated app built by a prompt doesn’t offer that.
Enterprises won’t vibe code their tech stack.
A global bank focuses on moving money, not maintaining a custom built, AI generated CRM. Vibe coding might make the initial build fast, but it doesn’t make maintenance free. Managing a fleet of custom, AI generated microservices is a nightmare that no CTO wants. Companies will always pay to outsource their non-core context (HR, CRM, Project Management) so they can focus on their core domain.
The disruptors themselves are seat-based SaaS businesses. The most glaring hole in the argument is that OpenAI and Anthropic, the companies supposedly killing the seat-based model, are seat based SaaS companies. ChatGPT Team and Claude for Business charge $25–$30 per user. They are selling a glorified assistant as a subscription. If the leaders of the AI revolution are leaning into the seat based model, the model isn’t dying – it’s being validated.
The value of SaaS has always been about outsourcing complexity for a predictable fee. AI doesn’t change that value proposition, it just changes the toolkit.