Labor economics & the AI-impact thesis
Revenue per employee and SG&A intensity across all 62 markets. Low rev/employee paired with high SG&A is where AI most directly attacks the fragmentation constraint.
The AI compression hypothesis
Internet and software markets have high revenue per employee because capital investment in distribution and product development produces leverage. Software: ~$600K per employee. Internet platforms: $1.5M–$2M+ per employee.
Cognitive services (legal, accounting, consulting, insurance, IT services) and physical services (trades, cleaning, landscaping) are stuck at $100K–$400K per employee because labor isthe product. These industries carry 30–75% SG&A ratios — a direct measure of how much cost is white-collar back-office work.
AI collapses the white-collar constraint. When routine analysis, quoting, scheduling, compliance, and underwriting are automated, the same dynamics that concentrated software (Aggregation Theory) should play out in these fragmented industries. The charts below rank markets by where that pressure has the most room to work.
Category averages
Markets ranked by leader SG&A % of revenue
Leader = the #1 player in each market. SG&A % is computed from the most recent 10-K income statement as SG&A / total revenue. Private firms (McKinsey, Kirkland & Ellis, Cargill, Stripe) use directional estimates. Click any bar for the market's full calculation.
Operating Overhead vs. Concentration
Each dot = one market. Y-axis: SG&A as % of revenue (top-3 revenue-weighted avg where researched, leader-only fallback otherwise). X-axis: CR3 = top-3 combined market share. Dot size scales with market size ($B). Click a dot for full market detail.
lib/labor.tsas a placeholder. Markets where the leader is private and no public filings disclose SG&A are excluded — see “no data yet” count above.Source Quality Hierarchy
Every market has a primary source plus 1–3 validation sources. Revenue numerators come from SEC 10-K filings; denominators validated against the Economic Census where available.
Data vintage: Q2 2025 research compilation. Labels updated 2026-Q2. Tier 1 sources refreshed from FY2025 SEC filings where available. For full methodology see the Methodology page.