Deploy Microsoft 365 Copilot connectors in the Microsoft 365 admin center

This article describes how to set up Microsoft 365 Copilot connectors in the Microsoft 365 admin center. The setup process is streamlined and requires minimal input, making it easy to create connections. You can also choose a custom setup to fine-tune specific settings.

Note

The setup process is similar for all Copilot connectors, with some differences. We recommend that you reference the deployment guide specific to your data source.

Prerequisites

Before you begin, make sure you have the following prerequisites:

• Admin access: You must be an AI administrator in the Microsoft 365 admin center.
• Data source credentials: Gather the required credentials and permissions for the data source.
• Service account (if applicable): Make sure that the service account has the necessary roles or permissions.

Deploy a connector

Use the following steps to deploy a connector:

1. Sign in to the Microsoft 365 admin center.
2. In the left pane, choose Copilot > Connectors.
3. On the Connectors tab, in the left pane, choose Gallery, and select the data source you want to connect (for example, ServiceNow Knowledge or Salesforce).
4. Enter a name to help users recognize the source in Copilot and search results. You can customize the default name.
5. Provide the URL of your data source. For example: https://your-organization-name.service-now.com.
6. Choose an authentication method to access the data source.
7. Deploy the connector to a subset of users for validation before a broader rollout.
8. Choose Create to deploy the connection. The connector begins indexing content using default settings.

Note

Most connectors use optimized default settings for access permissions, schema, and sync frequency. To customize the default settings, see Customize connector settings.

On the success screen, add a description that answers the following questions:

• What kind of content does this connection include?
• How do users refer to this content source?
• When do users access this content in their workflow?
• What are key characteristics of the content?

For guidance, see Enhance Copilot discovery of connector content.

Customize connector settings (optional)

You can customize the default values for connector settings. On the connector page in the admin center, choose Custom setup. This option includes three tabs:

• User
• Data
• Crawl

User settings

On the User tab, under Access Permissions, choose whether indexed data is visible to:

• Only users with access to the content
• Everyone in the organization

By default, users are mapped by matching their email to UserPrincipalName or Mail in Microsoft Entra ID. In the Map Identities section, you can provide a custom mapping formula.

Data settings

On the Data tab, under Manage properties, you can:

• Configure properties to be searchable, queryable, or refinable.
• Assign semantic labels and aliases to improve search relevance.
• Customize values for properties for certain connectors, like URL.

Content property

Select a Content property from the dropdown or use the default. This property supports full-text indexing, snippet generation, language detection, and relevance ranking.

You can use the system-generated ResultSnippet property in your result type to display dynamic snippets.

Aliases

In the Alias column, add aliases to normalize property names across multiple connections. This alias enables unified filters and queries.

Semantic labels

Assign semantic labels to source properties to integrate connector data into Microsoft 365 experiences. The following table lists the supported labels.

Label	Description
title	Title shown in search and other experiences
url	Target URL of the item
Created By	Creator of the item
Last modified by	Most recent editor
Authors	Collaborators
Created date time	Creation timestamp
Last modified date time	Last edit timestamp
File name	Name of the file
File extension	File type (PDF, DOC, etc.)

We recommend that you assign a property to the title label.

Incorrect mapping labels can affect search experiences. Not all labels need to have a property assigned.

Search schema attributes

You can set the search schema attributes to control the search functionality of each source property. A search schema helps determine what results are displayed on the search results page and what information end users can view and access.

Search schema attributes include options to Query, Search, Retrieve, and Refine. The following table lists the supported attributes.

Note

Properties with the int datatype can't be refined, even if marked as refinable.

Attribute	Function
SEARCH	Makes property content searchable
QUERY	Enables property-specific queries
RETRIEVE	Allows property to appear in search results
REFINE	Enables filtering on the search results page

Only string properties can be marked as searchable.

Crawl settings

On the Crawl tab, you can configure how often data syncs between the source and the connector index.

• Full crawl: Syncs all data at scheduled intervals.
• Incremental crawl: Syncs only new or changed data.

Note

Incremental crawls don't support permission updates. Run full crawls periodically to maintain sync accuracy.

Crawl scheduling

Configure crawl frequency and timing:

• Recurrence: Daily, weekly, biweekly, or monthly
• Days: Specific days of the week
• Frequency: Repeat interval (15 minutes to 12 hours)
• Start time: When the crawl begins
• Reset: Revert to default schedule

If fields are left blank, Copilot connectors choose optimal crawl times.

IP firewall rules

To secure access, configure your firewall to allow Copilot connector service IP ranges.

Region	Microsoft 365 Enterprise	Microsoft 365 Government
NAM	52.250.92.252/30, 52.224.250.216/30	52.245.230.216/30, 20.141.117.64/30
EUR	20.54.41.208/30, 51.105.159.88/30	NA
APC	52.139.188.212/30, 20.43.146.44/30	NA

Connector setting guidelines

Content property

We recommend that you select a Content property from the drop-down list, or keep the default selection if one is provided. The content property is used for full-text indexing and supports scenarios such as search result snippet generation, result cluster participation, language detection, HTML and text processing, ranking and relevance, and query formulation.

When you select a content property, you can use the system-generated ResultSnippet property when you create your result type. ResultSnippet acts as a placeholder for dynamic snippets that are generated at query time from the content property. When this property is included in a result type, snippets appear in search results automatically.

Aliases for source properties

You can add aliases to properties by using the Alias column. Aliases are friendly names that you can use in queries and when creating filters. They also help normalize properties from multiple connections by mapping different source properties to a single, common name. This approach allows you to create a single filter for a vertical that spans multiple connections. For more information, see Customize the search results page.

Semantic labels for source properties

You can assign semantic labels to source properties. Semantic labels are well-known tags provided by Microsoft that add meaning to your data and enable integration across Microsoft 365 experiences, including Copilot, enhanced search, people cards, intelligent discovery, and more.

The following table lists the supported semantic labels and their descriptions.

Label	Description
title	The title of the item shown in search and other Microsoft 365 experiences
url	The target URL of the item in the source system
Created By	The name of the person who created the item
Last modified by	The name of the person who most recently edited the item
Authors	The names of people who contributed to or collaborated on the item
Created date time	The date and time when the item was created
Last modified date time	The date and time when the item was most recently edited
File name	The name of the file
File extension	The file type, such as PDF or DOC

The properties on this page are preselected based on your data source, but you can change the selection if a different property is better suited for a specific label.

The title label is the most important semantic label. We strongly recommend that you map a property to this label so that your connection can participate in the result cluster experience.

Incorrect label mappings can degrade the search experience. Not all labels are required to have a property assigned.

Search schema attributes

Note

Properties with the int data type can’t be refined, even if they’re marked as refinable.

Search schema attributes control how each source property behaves in search. The search schema determines which results appear on the search results page and what information users can view and access.

Copilot connectors support the following search schema attributes.

Search schema attribute	Function	Example
SEARCH	Makes the text content of a property searchable and includes it in the full-text index.	If the title property is searchable, a query for Enterprise returns items that contain Enterprise in the title or body text.
QUERY	Allows querying against a specific property by name, either programmatically or by using query syntax.	If the Title property is queryable, the query Title:Enterprise is supported.
RETRIEVE	Allows a property to be displayed in search results and used in result types.	Only retrievable properties can appear in search results.
REFINE	Enables a property to be used as a filter on the Microsoft Search results page.	Users can filter by URL if the property is marked as refinable.

For all connectors except the File share Copilot connector, custom types must be configured manually. To enable search capabilities, the search schema must be mapped to a set of properties. The connection configuration assistant selects a default schema based on the source properties you choose, but you can modify it by selecting the appropriate attributes for each property on the search schema page.

Restrictions and recommendations for search schema settings

• The content property supports search only. It can’t be marked as retrieve or query.
• Rendering search results with the content property can cause performance issues. For example, using the Text field for a ServiceNow knowledge base article can significantly affect performance.
• Only properties marked as retrievable can appear in search results and be used to create modern result types (MRTs).
• Only string properties can be marked as searchable.
• All properties mapped to semantic labels must be retrievable.
• Properties with the int data type can’t be refined.
• A refinable property should also be queryable and retrievable.
• You can’t remove the retrievable attribute from a property.
• You can’t add or remove the refinable attribute from a property after setup.

Note

To update the schema after creating a connection, see Manage search schema.

Customize values for certain schema properties

Some connectors, such as ServiceNow, allow you to customize the values of specific schema properties, such as AccessURL, to meet organizational requirements. For example, if your organization uses a custom URL for ServiceNow knowledge articles instead of the default ServiceNow URL, you can configure the AccessURL property to point to the correct location. This customized URL is then used in Microsoft Search and Copilot responses.

Note

To customize schema property values, you must create a new Copilot connector connection. Editing an existing connection to change schema property values isn’t currently supported.

To customize the value of a schema property:

1. During connection setup, select Custom setup and authenticate your credentials.
2. Before publishing the connection, open the Content tab and go to Manage properties.
3. In the Properties table, select the property you want to customize (for example, URL).
4. In the side panel, locate the Default expression section.
5. Enter the expression in the New default expression box.
   • Reference variable components by using the exact property name enclosed in ${}.
6. Select Save changes at the bottom of the side panel.
7. Select Preview data at the top of the Content tab and scroll to the customized property to validate the expression.

Example

Assume that a ServiceNow knowledge article URL in your organization looks like this:

https://instancedomain.service-now.com/sp?id=kb_article&sys_id=1A2bcd3e45f6ghij17890kl1mn2345678o

Replace the variable component (sys_id) with the indexed property name (for example, SysId):

https://instancedomain.service-now.com/sp?id=kb_article&sys_id=${SysId}

The following table shows more examples.

Connector	URL in your organization	Default expression
ServiceNow Knowledge	https://contosos.service-now.com/serviceportal?id=kb_article_view&sysparm_article=KB0102362	https://contosos.service-now.com/serviceportal?id=kb_article_view&sysparm_article=${Number}
ServiceNow Knowledge	https://custom.contoso.com/knowp?id=kb_article&table=kb_knowledge&sys_id=a3f66636838e9a90	https://custom.contoso.com/knowp?id=kb_article&table=kb_knowledge&sys_id=${SysId}
ServiceNow Catalog	https://custom.contoso.com/sp?id=sc_cat_item&sys_id=bf21d3a54731465c34	https://custom.contoso.com/sp?id=sc_cat_item&sys_id=${SysId}
ServiceNow Catalog	https://contoso.service-now.com/1M_PorTAL?id=sc_cat_item&table=sc_cat_item&sys_id=4df777541b37ab4000	https://contoso.service-now.com/1M_PorTAL?id=sc_cat_item&table=sc_cat_item&sys_id=${SysId}
ServiceNow Ticket	https://contoso.service-now.com/esc?id=ticket&table=sc_req_item&sys_id=113851de1b7894&view=sp	https://contoso.service-now.com/esc?id=ticket&table=${EntityType}&sys_id=${SysId}&view=sp
ServiceNow Ticket	https://contoso.service-now.com/esc?id=ticket&sys_id=a3f66636838e9a9a&table=incident	https://contoso.service-now.com/esc?id=ticket&sys_id=${SysId}&table=${EntityType}

Add rules to customization for schema properties

For some connectors, you can override the default expression for specific items by adding rules based on property filters. Examples include:

• Knowledge Base for ServiceNow Knowledge
• Category for ServiceNow Catalog
• Entity type (such as incident or change_request) for ServiceNow Tickets

To add rules:

1. Complete steps 1–4 in Customize values for certain schema properties.
2. In the Set additional rules to configure expressions section, select Add new rule.
3. In the side panel:
   • Select the Filter property (for example, KbKnowledgeBase or Category).
   • Enter the Value using the exact case from your schema. Separate multiple values with commas.
   • Define the Expression to apply.
4. Select Save changes.
5. Save changes again at the bottom of the side panel.
6. Use Preview data on the Content tab to validate the customized expression.

Note

• If an item matches multiple rules, the first rule in the list is applied.
• URL changes take effect in Copilot results after the next scheduled full crawl.

Example

To apply a different URL expression for ServiceNow knowledge articles in the IT and HR Policy knowledge bases, set KbKnowledgeBase as the filter, specify IT, HR Policy as the value, and use an expression such as:

https://instancedomain.service-now.com/esc?id=kb_article_view&sysparm_article=${Number}

Guidelines for crawl settings

The refresh interval controls how often data is synchronized between the source system and Microsoft Search. Optimal schedules vary by data source and depend on how frequently content and metadata change.

Two refresh types are supported: Full refresh and Incremental refresh. Incremental refresh isn’t available for all data sources.

A full refresh processes and indexes all changed items, regardless of previous crawls. Use a full refresh in scenarios such as:

• Detecting deleted items.
• Recovering from incremental crawl failures.
• Updating ACLs.
• Modifying crawl rules.
• Updating the connection schema.

An incremental refresh processes only items that were created or modified since the last successful crawl. This approach is faster and is best suited for frequent content or metadata updates.

Note

Incremental crawls don’t currently process updates to permissions.

Even when using incremental refreshes, you should run full refreshes periodically to ensure data consistency between the source system and the search index.

Crawl scheduling

You can configure full and incremental crawls by using the advanced scheduling options on the Refresh settings page. Some connectors don’t support incremental crawls. For others, incremental crawling is optional and enabled by default.

A default crawl schedule is selected based on the connector type, but you can change it during connection creation or by editing the connection after publishing.

Available options include:

• Recurrence: Daily, weekly, biweekly, or every four weeks.
• Day: Run crawls on specific days of the week.
• Run once a day: Run a single crawl at a specified start time.
• Frequency: Repeat crawls within a day. Intervals range from 15 minutes to 12 hours.
• Starting time: The time when the crawl begins.
• Reset: Restore the default schedule for the connector.

Also consider the following:

• If fields are left blank, Copilot connectors automatically select an optimal start time.
• System load can delay crawl start times by up to an hour.
• If a crawl overlaps with the next scheduled crawl, the next crawl runs only if it’s a different type (full or incremental).

IP firewall rules

If your data source uses IP firewall rules, allow access to the Copilot connectors service by permitting the following IP ranges.

Region	Microsoft 365 Enterprise	Microsoft 365 Government
NAM	52.250.92.252/30, 52.224.250.216/30	52.245.230.216/30, 20.141.117.64/30
EUR	20.54.41.208/30, 51.105.159.88/30	N/A
APC	52.139.188.212/30, 20.43.146.44/30	N/A

Related content

• Prerequisites
• Connectors gallery