The ARR era is over: AI collapses SaaS multiples from 114x to 4x
akumar · x · 2026-10-14
A long-form history of the subscription model's rise and fall: software was capex from 1969-1999 (perpetual licenses — Oracle, SAP, Microsoft); Salesforce launched at $50/user/month in 1999, and after its 2004 IPO ARR was born as a valuation tool. Adobe killed perpetual licenses in 2013 — revenue dropped 8%, then grew from $4.4B to $15B, and the industry copied the playbook.
2021 was the peak: median late-stage valuations hit 114x ARR (vs 15x four years earlier); Snowflake traded at 80-100x. The author argues the whole financial stack assumed customers could never write the software themselves — a decade of accumulated code was an unreplicable moat. AI broke that assumption.
In 2026 software trades at a P/E discount to the S&P 500 for the first time, $2T erased, private SaaS down from 100x+ to 4-8x, a16z publicly tells founders to stop stressing ARR, and EV/EBITDA becomes the new standard. Founders and investors still optimizing for ARR are using a measuring stick from a world that no longer exists.
More from AGI Musings
- MIT's Rakhlin: universities must prepare for AI outsmarting their researchers — nordicinst · 2026-10-14
- Faggella: don't face AGI flat-footed — keep agency, use AI, stay contributive — danfaggella · 2026-10-14
- Michael I. Jordan: superintelligence is science fiction, doomsday talk demoralizes youth — BachFrancis · 2026-10-14
- Anthropic co-founder Jack Clark moves recursive AI self-improvement forecast to 2027 — 141_1337 · 2026-10-14
- Paul Graham: nearly all standard startup advice still applies in the AI age — ycombinator · 2026-10-14
- "Software is dead" debunked: vibe-coding on the fly costs tokens, compute and 15-minute waits — bendee983 · 2026-10-14