AI labs consistently spend 5-10% of revenue on expert data, Mercury CEO says
StephanSturges · x · 2026-10-10
Mercury AI CEO Brendan Foody noted that the average lab or AI-native company spends 5-10% of revenue on expert data, a figure that is surprisingly consistent across the industry.
Reposter Stephan Sturges commented that this ratio is "going to go waaay up," suggesting demand for expert data — domain expert labeling and demonstrations used for model training — will keep pushing that cost share higher.
More from Venture
- OpenAI's Thibault Sottiaux on how useful ChatGPT plugins can become a business — lennysan · 2026-10-10
- AI inference could hit $350B next year, pushing software margins from 72% down to 30-50% — Sethwinterroth · 2026-10-10
- Nebius vs CoreWeave: net debt 0.2x vs 2.9x revenue explains the stock divergence — Beth_Kindig · 2026-10-10
- Ramp data: 80% of OpenAI and Anthropic enterprise revenue comes from just 1% of customers — iamtrask · 2026-10-10
- A new app pulled the most aggressive cold-start growth hack via email access — msg · 2026-10-10
- $55 billion is equal to a 1% swing of NVIDIA stock — JosephJacks_ · 2026-10-10