Start treating AI agents like first-class users
Blog post from Mixpanel
Automated web traffic reportedly surpassed half of all requests by mid-2026, with a rapidly growing portion coming from AI agents that act on users’ behalf to research, compare, book, buy, and manage software tasks. Unlike traditional crawlers or malicious bots, these agents carry human intent but behave very differently from people, generating much higher event volumes, favoring structured and efficient information, and potentially distorting conventional product metrics such as time on site, cart activity, and clicks. The piece argues that companies should treat agents as first-class users by creating agent-friendly experiences through APIs, MCP servers, and structured data, while maintaining visibility into interactions that combine human and agent activity. It describes Mixpanel’s planned Agent Classification capability as a way to distinguish human, bot, and agent traffic in analytics, enabling teams to analyze agent funnels, session behavior, experiments, retention, and conversion separately and alongside human journeys.
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.