A marketplace of digital AI agents of real people

TAM TOP-DOWN

Grand View Research and Precedence Research both size the global AI agents market at roughly $7.6B in 2025, covering software, platforms, and professional services. A marketplace specifically for persona-based digital replicas of real people is a subset of the discovery and licensing layer. Marketplace/platform intermediation historically captures 20-30% of a software market's gross spend; applying the midpoint of 24% to $7.6B yields the TAM. This is conservative because it excludes embedded agentic spend in CRM and ERP.

$7.6B x 24% marketplace layer = $1.8B TAM

TAM BOTTOM-UP

31% of enterprises run at least one AI agent in production (McKinsey/S&P Global). The global enterprise software market is $291.75B in 2025; assuming 50,000 large enterprises as the relevant universe (consistent with Fortune 500 scale and tier-1 global firms), and that 5% would trial a real-person agent at $500/month average, this cross-checks the top-down. Result is $1.5B, broadly consistent with the top-down $1.8B, so no divergence flag is triggered at the TAM level.

50,000 enterprises x 5% trial rate x $500/month x 12 = $1.5B

SAM FILTERS

Filter 1: Geography. North America ($2.45B) plus Europe ($2.31B) represent the primary markets for compliant, licensed digital-identity products; combined regional share is roughly 63% of global TAM, yielding $1.13B. Filter 2: Real-person agent niche. No direct comp exists; assuming 16% of agent platform spend could flow to persona-licensed agents (analogous to celebrity/influencer licensing economics) yields $180M top-down SAM. Bottom-up: 3,000 licensable agents at $150/month x 12 months x 10% marketplace take-rate on a 100-buyer base yields $54M. Divergence exceeds 2x, flagged.

Top-down: $1.13B x 16% = $180M | Bottom-up: 3,000 agents x 100 buyers x $150/mo x 12 x 10% = $54M

SOM BUILD

Using the conservative bottom-up SAM of $54M as the base. A competent new entrant in year 1 realistically signs 200 licensed agents and 20 enterprise buyers, generating roughly $360K. By year 3, scaling to 1,800 active agent listings across 150 buyers at $250/month average list price, with a 10% marketplace rake, yields $5.4M. This is 3% of the bottom-up SAM and reflects the slow consent-acquisition cycle, incumbent competition from Salesforce AgentExchange and Kore.ai, and the 40%-plus project cancellation headwind cited by Gartner.

1,800 agents x 150 buyers x $250/mo x 12 x 10% rake = $5.4M year-3 SOM