Particle.news

After SaaS Selloff, AI Agents Force Repricing as Capital Shifts to Moat-Heavy, AI‑Native Plays

Investors now favor AI-native software built on proprietary data with deep workflow control.

Overview

  • A February selloff erased nearly $1 trillion from software and services stocks, and subsequent AI product launches like Anthropic’s Claude Code and Cowork continued to rattle sector ETFs and incumbents’ shares.
  • Top VCs and incumbents reject a 'SaaSpocalypse' narrative, with Sequoia’s Alfred Lin and Snowflake’s Sridhar Ramaswamy arguing durable value will persist where data integrity, governance and trust are embedded.
  • Pricing models are under pressure as per-seat economics weaken when agents do multi-seat work, pushing SaaS toward consumption- and outcome-based approaches, exemplified by Sierra reaching $100 million ARR with outcome pricing.
  • Investor focus is moving away from thin workflow layers, generic horizontal tools and easily replicated interfaces, and toward AI-native infrastructure, proprietary data moats and products that own mission-critical workflows.
  • The near-term IPO window for venture-backed SaaS has cooled, with a Crunchbase report expecting no filings, while investors and strategics signal more M&A as incumbents seek AI capabilities and startups eye legacy user bases.