AI integration for SMEs

Positive has introduced the Model Context Protocol (MCP) to allow AI assistants to connect directly to the marketing and commercial campaigns of small and medium-sized enterprises (SMEs). This system enables a single AI agent to manage the entire process, from visibility in search engines to sales, without switching tools or breaking the authorization chain.

Simplified marketing workflows

The goal of MCP is to let SMEs link their preferred AI assistant to their entire marketing and sales process without using multiple tools. Traditionally, marketing software required users to open interfaces, navigate menus, and switch between tools. MCP changes this by allowing AI agents to handle these tasks seamlessly.

Access within existing AI environments

A team already working in platforms like Claude, ChatGPT, or Copilot can now directly upload data and features from Positive into those environments. Dashboards remain accessible through Positive interfaces, but MCP adds a new way to interact: through the AI agent. Positive’s native AI, Uma, is designed to manage the different products in the suite.

Real-world marketing with AI

Imagine a marketing manager asking her AI assistant, "Which pages of our site have lost visibility this week?" Positive Surfer would analyze the data and report the results directly. She could then ask the AI to prepare LinkedIn posts or identify engaged contacts, all without switching tools. In the future, Positive User will also allow such actions while respecting user rights.

Streamlined marketing processes

MCP eliminates the need to switch between tools, copy data, or rebuild context at each step. A single AI agent can manage a campaign from start to finish, making the Full Loop — the entire marketing and sales journey — controllable by AI. This loop includes visibility, traffic, content creation, engagement, and data collection, all of which can now be managed by an AI agent.

Headless architecture for AI

Thomas Thelliez, CTO of Positive, explains that MCP enables any AI assistant to manage the Full Loop from start to finish. This represents a shift from a connected suite to a headless architecture, where AI can operate without needing to access traditional interfaces.

Growing AI adoption

AI agents are moving from testing to real-world use. According to McKinsey, 40% of large companies are now deploying AI agents on a large scale, up from 27% the previous year. MCP adoption is also growing rapidly, with SDK downloads increasing from 97 million in December 2025 to nearly half a billion per month in July 2026.

SMEs benefit from simplified integration

SMEs often lack the resources to build and maintain multiple integrations between tools, data, and AI models. MCP reduces this complexity by allowing a single agent to interact with various stages of marketing and sales without needing separate integrations. Positive remains compatible with any AI assistant, whether it's Uma or a third-party tool.

Gradual rollout with security

Not all features will be available to AI agents immediately. Positive Surfer, Iconosquare, and Signitic have been released with MCP servers, allowing specific tasks like SEO audits, social media scheduling, and brand compliance checks. Positive User will be added later, with read-only access first, followed by write actions once security standards are fully met.

No extra cost for access

Access to MCP is included in current Positive packages without new subscriptions, licenses, or passwords. Existing users can gradually integrate their workflows with AI, while new users can start from AI platforms like Claude or ChatGPT. AI agents can only access data and actions that the user is already authorized to use, ensuring security remains a priority.