A chat agent that talks to everyone the same way feels like a machine. A chat agent that knows your name, remembers your account, and skips the questions it already has answers to feels like real help. On a Power Pages site, you already hold a lot of information about your signed-in users in their contact record. With a little setup, your Copilot Studio agent can read those contact fields and use them to make every reply feel personal. In this guide I will show you how.
Why Personalization Is Worth It
Think about how it feels when a support chat opens with “Hi Sarah, I can see your recent order” instead of “Please enter your name and order number.” The first one feels warm and saves time. The second one feels cold and makes the visitor repeat things you already know. Personalization is not just a nice touch. It removes friction and builds trust.
When the agent already knows who it is talking to, it can skip past the basic questions and get straight to helping. That makes the whole conversation shorter and more pleasant for everyone.
The Key Idea: Pass the Contact Details to the Agent
Here is the core of how this works. When a signed-in user opens the chat, Power Pages knows who they are because they are logged in. The contact record holds their name, email, and whatever other fields you store, like their company or account type. The trick is to pass those details from the page into the agent at the start of the conversation. Once the agent has them, it can use them in its replies.
So the flow is simple. The page reads the contact fields, hands them to the agent, and the agent uses them. You are bridging what the site knows with what the agent can say.
Step One: Set Up Variables in the Agent
In Copilot Studio, you prepare the agent to receive these details. In the Conversation Start system topic, you add a Question node but leave the question itself blank. You are not really asking the visitor anything. You are using the node as a container that can receive a value from outside the chat. You set the variable to be global so any topic can use it, and you mark it so that external sources are allowed to set its value.
You repeat this for each contact field you want, like first name or account type. Give each variable a clear name so you can use it easily later. These variables are the slots the agent will fill with the contact details.
Step Two: Send the Contact Fields From Power Pages
On the Power Pages side, you read the signed-in user’s contact fields and pass them into the agent when the chat loads. Power Pages can read these fields through Liquid, since it already knows the logged-in user. You then hand those values to the agent as the context variables you set up. Now, the moment the chat opens, the agent already holds the person’s name and any other detail you sent.
Because the user is signed in, this happens quietly in the background. The visitor does not type anything. They just open the chat and the agent already knows them.
Step Three: Use the Fields in Your Replies
Now the fun part. With the contact details sitting in variables, you can weave them into the agent’s messages. The greeting can use the first name. A topic about orders can use the account type to decide which path to take. A support topic can skip asking for an email because the agent already has it.
You can also use these fields to branch the conversation. For example, a premium customer might be offered a different option than a standard one. Because the agent knows the account type up front, it can tailor the whole experience without asking a single extra question.
Keep Privacy in Mind
Using personal details means you must handle them with care. Only pass the fields the agent truly needs. There is no reason to send sensitive information that the conversation will never use. The less you pass, the safer you are.
Make sure this personalization only runs for signed-in users, since anonymous visitors have no contact record to read. And respect your table permissions, so the agent only ever works with data the user is allowed to see. Personalization should make people feel known, not watched, so keep it focused and respectful.
Test With Real Accounts
To check your work, sign in as a test user who has a full contact record and open the chat. Confirm the agent greets you by name and uses the other fields the way you planned. Then sign in as a different user to make sure the agent picks up their details, not the first user’s. This catches mix-ups where the wrong values carry over.
Also test as a signed-out visitor. The agent should still work, just without the personal touch, and it should never show a blank where a name should be. A small fallback, like “there” instead of a missing name, keeps things smooth.
Wrapping Up
Personalizing a Copilot Studio agent with Power Pages contact fields turns a generic chat into a warm, helpful conversation. You set up variables in the agent, pass the signed-in user’s contact details from the page, and use those details to greet people by name and skip questions you already have answers to. Keep the data minimal and protected, test with real accounts, and your agent will feel less like a bot and more like someone who already knows your customer.