Blab AI
AI-Native Startups Operate Leaner, Raise More Capital Per Employee, Study Finds
When venture capitalists hunt for the next high‑growth company, they’re finding a new model that relies on artificial‑intelligence (AI) to scale without the bulk of a traditional team.
A joint study by Harvard Business School and INSEAD, which examined almost 50,000 venture‑backed firms launched between 2020 and 2024, shows that AI‑native startups are smaller, flatter and raise more money per employee than comparable non‑AI peers.
The researchers drew data from Y Combinator and PitchBook, identifying AI‑first companies through Y Combinator’s self‑tagging system and then comparing their organizational metrics to those of traditional startups. The key findings are:
AI‑native firms average 25 % fewer employees, hiring 15 % fewer entry‑level workers and 15 % fewer managers. Their hierarchies are flatter, with a higher proportion of senior staff and a lower share of junior employees. Engineers make up 13 % more of the workforce in AI firms than in non‑AI startups. AI companies emphasize engineering roles while de‑prioritizing sales, finance, operations and administration. Despite their leaner size, AI startups secure similar total funding amounts and achieve valuations comparable to their non‑AI counterparts. On a per‑employee basis, AI firms raise roughly 20 % more capital and command higher valuations per employee.
The study also documents a surge in AI‑first deals at Y Combinator. In 2024, the number of AI‑first startups in the accelerator’s cohort was nearly eight times the average for 2020.
"Embedding AI into products, rather than simply adding AI tools to existing workflows, is a primary way startups scale knowledge work without large teams," the authors wrote.
The authors note that AI firms tend to cluster in Silicon Valley, with workforces that are more likely to be male, hold advanced degrees and come from prestigious employers and institutions.
Rather than pointing to a reduction in jobs, the study suggests that AI‑native startups achieve similar market impact with fewer people because AI technologies automate tasks that traditionally required larger teams.
The findings carry implications for investors, founders and policymakers. Investors may view the higher capital‑per‑employee metric as a sign of efficiency. Founders are reminded of the potential to build AI‑centric products with lean teams. Policymakers may use the concentration of AI talent in Silicon Valley and the demographic profile of these teams to inform workforce development and diversity initiatives.
Hyunjin Kim of INSEAD and Rembrand Koning of Harvard Business School, who authored the study, did not forecast future trends; they simply documented the current state of AI‑native startups.
In short, AI‑native startups are smaller, flatter and raise more capital per employee than their non‑AI counterparts, while maintaining comparable valuations and funding levels—a shift toward leaner organizational structures that leverage AI to scale product development and operations.
A joint study by Harvard Business School and INSEAD, which examined almost 50,000 venture‑backed firms launched between 2020 and 2024, shows that AI‑native startups are smaller, flatter and raise more money per employee than comparable non‑AI peers.
The researchers drew data from Y Combinator and PitchBook, identifying AI‑first companies through Y Combinator’s self‑tagging system and then comparing their organizational metrics to those of traditional startups. The key findings are:
AI‑native firms average 25 % fewer employees, hiring 15 % fewer entry‑level workers and 15 % fewer managers. Their hierarchies are flatter, with a higher proportion of senior staff and a lower share of junior employees. Engineers make up 13 % more of the workforce in AI firms than in non‑AI startups. AI companies emphasize engineering roles while de‑prioritizing sales, finance, operations and administration. Despite their leaner size, AI startups secure similar total funding amounts and achieve valuations comparable to their non‑AI counterparts. On a per‑employee basis, AI firms raise roughly 20 % more capital and command higher valuations per employee.
The study also documents a surge in AI‑first deals at Y Combinator. In 2024, the number of AI‑first startups in the accelerator’s cohort was nearly eight times the average for 2020.
"Embedding AI into products, rather than simply adding AI tools to existing workflows, is a primary way startups scale knowledge work without large teams," the authors wrote.
The authors note that AI firms tend to cluster in Silicon Valley, with workforces that are more likely to be male, hold advanced degrees and come from prestigious employers and institutions.
Rather than pointing to a reduction in jobs, the study suggests that AI‑native startups achieve similar market impact with fewer people because AI technologies automate tasks that traditionally required larger teams.
The findings carry implications for investors, founders and policymakers. Investors may view the higher capital‑per‑employee metric as a sign of efficiency. Founders are reminded of the potential to build AI‑centric products with lean teams. Policymakers may use the concentration of AI talent in Silicon Valley and the demographic profile of these teams to inform workforce development and diversity initiatives.
Hyunjin Kim of INSEAD and Rembrand Koning of Harvard Business School, who authored the study, did not forecast future trends; they simply documented the current state of AI‑native startups.
In short, AI‑native startups are smaller, flatter and raise more capital per employee than their non‑AI counterparts, while maintaining comparable valuations and funding levels—a shift toward leaner organizational structures that leverage AI to scale product development and operations.
