Agent-Friendly APIs Are the First Step to Autonomous Revenue

Agent-Friendly APIs Are the First Step to Autonomous Revenue

5d ago
SaaStr ARMARM Gtm_strategy

The Gist

  • Automated bot traffic surpassed human traffic for the first time in 2024
  • RAG-based agent traffic grew 49% in early 2025 and continues to accelerate
  • APIs must shift from human-first to agent-first design to reduce ambiguity and latency
Key Quotes

AI Agents will have no use for many pre-AI products like Marketo, Gong, Salesloft, Outreach, Atlassian, Monday and more. They'll just bypass them and do a better job ... themselves. When the agents pick.

Jason Lemkin

The first 'developer' evaluating your product isn’t a human. It’s an agent, on behalf of a human, deciding whether your API is worth integrating against or whether the competitor one tab over is easier.

Jason Lemkin
Key Insights
  • Agent-friendly APIs are now essential for B2B companies, as AI agents have become the primary consumers of APIs, surpassing human developers.
  • Traditional APIs designed for human developers are insufficient for AI agents, which require explicit, unambiguous, and self-correcting interfaces.
  • Companies that fail to adapt their APIs for AI agents risk losing customers to competitors with more agent-friendly solutions.
  • Salesforce is a leading example of a company successfully adapting its API for AI agents, while Marketo is failing to meet modern standards.
  • MCP (Machine-Client Protocol) servers and agent SDKs are now critical components for making APIs accessible to AI agents.
  • Agent-first error responses, which include structured hints for self-correction, can reduce wasted tokens by 60% or more.
Actionable Takeaways
  • Implement an MCP server and agent SDKs to make your API accessible to AI agents.
  • Redesign error responses to include structured, machine-readable hints for self-correction.
  • Audit and update your API documentation to ensure it is agent-friendly, with clear instructions for use and error handling.
  • Prioritize idempotency and webhooks in your API design to reduce agent retries and improve reliability.
Data Points
  • 49% (Growth in RAG-based agent traffic in early 2025.)
  • 60%+ (Reduction in wasted tokens after implementing agent-first error responses.)
  • 75% (Production APIs with endpoints that don’t match their OpenAPI spec.)
  • $60K/year (Cost of Marketo, which fails to meet agent-friendly API standards.)
  • 5,000+ (MCP servers in production as of 2026.)
  • 844,000+ (Sites that have shipped llms.txt as of 2026.)

RevBots.ai View:

Companies that don’t optimize their APIs for AI agents will bleed churn as competitors pull ahead.

Full Story: SaaStr →