This chart and analysis reveals how Model Context Protocol (MCP) usage has concentrated in development infrastructure, with 54.4% of all sessions focused on technical tasks. MCP is an open standard that lets AI assistants connect to external tools, essentially allowing AI to read from and interact with other systems. Whilst Claude pioneered MCP support, adoption has remained largely technical, as this chart clearly shows with minimal usage in consumer categories like personal management.
ChatGPT’s latest (beta) update could change this trajectory. OpenAI just enabled full MCP write access for Plus and Pro users, meaning ChatGPT can now actively modify external systems, not just read from them. This transforms ChatGPT from a query interface into an active automation platform.
Now, with ChatGPT’s mainstream reach, MCP could expand beyond developer tools, but only if it learns the right lessons. As Vercel’s team discovered, early MCP tools failed because they mirrored traditional software integrations/APIs, forcing LLMs to orchestrate multiple steps. Much better results can be achieved with workflow tools that handle complete intentions in single operations. This shift is exactly what consumers need: instead of learning to chain commands (“create document, then share it, then notify team”), they can complete natural goals. When MCP tools match how people and their AIs think rather than how APIs work, the protocol will move on from being a technical curiosity to part of our everyday lives.