Who:
- Jason Lemkin and Rory O'Driscoll of SaaStr: two of the most influential voices in SaaS economics
What Happened:
- SaaStr publicly predicted AI tokens will replace 25% of GTM headcount by 2027, with Uber already capping engineer token spend at $1,500/month.
- They revealed startups like McCor now spend more on tokens than engineering salaries, with 80 engineers.
- Declared the historical 13% tool spend ceiling for EDA software will shatter as AI workflows dominate.
Why It Matters:
- Forces every CFO to recalculate the breakeven point between human labor and AI token costs.
- Marginal roles (QA, inbound SDRs, low-tier CS) face immediate existential risk as token efficiency improves.
- Validates the ARM thesis that AI-native workflows will compress traditional GTM hierarchies.
ARM Impact:
- Tab Hopper (Stage 1 (Tab Hopper)): Manual roles will be first to face token replacement pressure.
- AI Sprinkler (Stage 3 (AI Sprinkler)): Token spend becomes a KPI alongside headcount for engineering teams.
- ARM (Stage 4 (Autonomous Revenue Master)): Companies like McCor prove fully token-weighted org structures can scale.
What to Watch:
- Monitor Uber's $1,500/month engineer token cap: if it holds, it sets the market benchmark.
- Track public SaaS earnings for mentions of "token efficiency" as the new gross margin metric.
- Watch for the first Fortune 500 to hit 20% token-to-headcount spend ratio (likely in dev tools first).