Dynamics 365 Customer Insights – Data 2026 Release Wave 1 Feature: Smarter Agents
Organizations are rolling out agents to help users work through data in Dynamics 365 systems, but they are only as useful as the data they can see. Smarter agents need better customer context, and that is the focus of this 2026 Release Wave 1 feature.
Today, many agents are limited by the systems they are directly connected to. A Copilot Studio agent might see a lead or contact record in Dataverse, but it may not automatically understand the full customer relationship behind that record. Purchase history, loyalty status, segment membership, churn risk, and Customer Lifetime Value may live somewhere else, unless a team has manually exposed that data through connectors, flows, APIs, or custom integrations.
This release is not about creating new customer intelligence. It is about making existing Customer Insights data easier for agents to discover, understand, and use inside Copilot Studio. Public preview is currently planned for June 2026, though Microsoft notes that release-plan functionality and dates can change. There is no General Availability date listed for this update.
What Makes This New?
The important change is not that Customer Insights suddenly has new analytics. Many of the underlying capabilities already exist: unified profiles, calculated measures, segments, predictive scores, and churn indicators. These insights should become much easier for agents to use.
MCP, or Model Context Protocol, allows Copilot Studio agents to connect to servers that expose tools and resources. Instead of building every path between an agent and customer intelligence, organizations can give the agent a more direct way to discover and use the right Customer Insights data.
Why It Matters
Imagine a seller looking at a lead who recently attended a webinar. In CRM, that person might look like a basic lead. But Customer Insights may know much more. The person could already be a paying customer, have multiple contact records, own several products, belong to high-value segments, hold rewards points, and show elevated churn risk.
That difference matters. Treating that person like a cold lead would be a mistake. Treating them like a known, valuable customer creates a very different conversation.
The Real Shift
Today, Copilot Studio requires additional setup: connections into Dataverse, connector actions, and possibly flows, APIs, or custom code to access Customer Insights data.
In the future, the setup should be less about manually wiring every insight through connectors, flows, APIs, or custom code, and more about letting the agent discover approved Customer Insights resources through the MCP server. Ideally, this all leads to more accurate answers, more personalized customer engagement, and higher conversion rates and overall customer satisfaction.
For sellers, this means fewer ācold leadā mistakes. For service teams, it means the agent can see whether someone is a high-value customer before giving a generic answer. For marketers, it means journeys can be informed by the customerās actual relationship, not just the latest form fill.
