What's my plan?

Available on all Suite plans

Tip: End-user authentication for messaging is only available for the Web Widget and mobile SDK channels.

Authenticating end users is important for many business purposes. When providing conversational support, users might try to carry on a conversation between devices and channels. By authenticating the end users, you can make sure all points of contact are associated with the correct end user. This can enhance the quality of support your agents provide and increase the security of sensitive information that might come up while agents assist your end users.

This article contains the following topics:
  • About end-user authentication for messaging
  • Creating and sharing a signing key
  • Authenticating end users with only an external ID
  • Incorporating email identities into your end user authentication

About end-user authentication for messaging

There are many reasons to authenticate end users for messaging, including:
  • Increasing confidence that the end users your agents are talking to are who they claim to be.
  • Making it possible to identify a single user across channels if you embed the widget on multiple domains or link to external services, such as Shopify.
  • Making it possible to identify a single user across different devices and browsers.

Terminology for messaging authentication

Before you get started, though, it's important to understand the following terminology:
  • JWT: Zendesk uses signed JSON Web Tokens (JWTs) to authenticate end users for messaging. These tokens contain details that verify the identity of the end users. For more information about JWT, see jwt.io.
  • Signing key: A signing key is created by a Zendesk admin in Admin Center and shared with a developer on your team, who then uses it to sign the JWT as necessary.
  • External ID: An alphanumeric string, such as a user name or ID from an external system, that is unique to each user. This is the primary identification for messaging authentication, even when an email address is included in the JWT.
  • User name: (Optional) The name of the end user associated with the external ID or email address. If you include the user's name in the JWT, it appears in the Agent Workspace. This information can help agents communicate with end users.
  • Email: (Optional) The unique email address associated with an end user.

Overview of implementing messaging authentication for end users

Implementing end-user authentication for messaging requires collaboration between admins and a developer on your team. Zendesk uses JSON Web Tokens (JWTs) to authenticate messaging end users.
  1. The process of messaging authentication begins with an admin generating a signing key and providing it to a developer. Then the developer uses the signing key to implement a back-end service that can create signed JWTs for users as requested.
  2. When requested, the back-end service creates and returns signed JWTs to your website or mobile app. The JWTs created by this service must include a unique external ID and, optionally, an email address to identify the end user.
  3. Any time the user is logged in, your website or app needs to call an equivalent login API available for web widget and mobile SDKs, at which time the JWT is passed to Zendesk to verify the claimed identity of the user.

Requirements for authenticating end users for messaging

The following must be true to authenticate messaging end users:
  • Zendesk Agent Workspace is activated.
  • You're using the web widget or mobile SDK channels for messaging.
If you're using email identifies to authenticate messaging end users, the following must also be true:
  • You associate email identities with end users.
  • If you want authenticated users to have verified email identities, the JWTs you issue to end users must contain both the email and email_verified: true claims.

Agent experience when authenticating messaging end users

Agents see a green check mark icon next to the visitor's name and the end user's external ID is visible next to the user's profile when the end user is authenticated.

Additionally, each response posted by an end user after they are authenticated has the check mark icon.

Authenticated end user can participate synchronously in a conversation across devices. For unauthenticated end users, separate user records and conversations are created for each device they use. If an end user authenticates mid-conversation, the conversations pre- and post-authentication are automatically merged to provide continuity for the agent and end user.

If you're using email identities, the mapping of end users to user records varies based on your email identity settings. With most configurations, imposters appear as separate user records and won't have the green check mark next to their name because they can't be successfully authenticated. If you permit unauthenticated users to claim verified email addresses, email identities are associated with the first user who claims the address. If two end users claim the same address, the email identity is attached to the user record for the end user who is able to authenticate and verify the email address.

In the event of a conflict, verified email identities supersede unverified email identities. For example, if an unauthenticated user provides an email address that is already associated with a verified identity, Zendesk creates a new unauthenticated session and user record without an email identity for the unauthenticated user who is attempting to impersonate a verified user. Similarly, if an unauthenticated user provides an email address but can't verify it, but later another user provides the email address and is able to verify it and sign in with their JWT, the email identity will be attached to the authenticated user and removed from the unauthenticated user's record.

Agents have the ability to merge duplicate user records and manually add an email identity to user records. We recommend training your agents to perform identity verification checks with end users before taking either of these actions.

End user experience with authentication for messaging

After you implement end-user authentication for messaging, end users shouldn't notice much of a difference. After they have been authenticated and their identity is verified with Zendesk, end users aren't prompted to provide their name or email address by messaging bots as part of the default messaging response.

Conversations for authenticated end users are synced across devices when the end user is authenticated. Separate user records and conversations are created for unauthenticated end users. If an end user authenticates mid-conversation, the anonymous conversation that was created before they signed in is automatically merged with the authenticated conversation to provide conversational continuity.

Creating and sharing a signing key

Signing keys are used by developers to create JWTs for end users. You must be an admin to create a signing key. You can create a maximum of 10 keys. If you attempt to create a new key after reaching your limit of 10, you are prompted to delete unused keys.

To create and share a signing key
  1. In Admin Center, click Account in the sidebar, then select Security > End user authentication.
  2. Click the Messaging tab, then click Create key.

    If you are creating your first key, Create key appears at the bottom of the page. Otherwise, it appears in the top-right corner.

  3. Enter a Name for the key and click Next.
  4. When prompted, click Copy to copy the shared secret and then click Hide key forever.

    The key is saved and an ID is automatically assigned. You can find a key's ID in the list of keys on the Messaging tab of the End user authentication page.

  5. Confidentially send the key's ID and the shared secret you copied to your developer.

Authenticating end users with only an external ID

After an admin creates a signing key, the developer on your team can use it to create a service that generates JWTs for end users. When your developer creates JWTs using an external ID, the payload needs to include the following information:
  • external_id: (Required) This is the unique alphanumeric string that can be used to identify each user. Commonly, these IDs are pulled from external systems. An ID can't exceed 255 characters.

    Zendesk uses the external_id as the primary identifier for users authenticating over messaging. When authenticating a user with a valid JWT, Zendesk first resolves an existing user with the external_id. If no external_id match is found, users are resolved using the email address provided. See Issuing JWTs with email addresses.

  • scope: (Required) The caller's scope of access. The only valid value is user.
  • name: (Optional) The name of the user. Including the name in the JWT payload enables Zendesk to display the user's name in the Agent Workspace and helps your agents provide more customized support.
For example: 
{
    "external_id": "12345678",
    "scope": "user",
    "name": "Jane Soap"
}

Incorporating email identities into your end user authentication

In addition to the external ID, you can choose to use email identities with authenticated users, unauthenticated users, or both:
  • Authenticated users are authenticated through signed JWTs.

    The use of JWTs provides a trustworthy approach because the content of a signed JWT can't be tampered with by end users. If you're concerned about impersonation attacks, you should restrict email identities to authenticated end users. This is the most secure option and is the default approach for new Zendesk accounts.

  • Unauthenticated users are end users who provide an email address in response to a prompt by a Zendesk bot.

    Keep in mind that permitting the use of email identities for unauthenticated users can make you vulnerable to people impersonating other users by providing an email address they don't own.

Configuring email identities

On new Zendesk accounts, email identities are on and configured to use only verified email addresses. This is the most secure option. Older accounts are configured to use both verified and unverified email addresses. Before adding email addresses into your JWTs, you should review the options and update your settings as needed.

To turn on email identities
  1. In Admin Center, click Channels in the sidebar, then select Messaging and social > Messaging.
  2. Click Manage settings.
  3. Click Email identities, and then select one of the following options:
    • Use only verified emails: Email identities are created only for users who are authenticated and have a verified email address included in their issued JWT. Agents can still manually add an email identity to a user record.

      With this option, agents see the email address provided by unauthenticated end users in the chat history, but they won't see an email identity attached to the user in the essentials card. If an agent needs to follow up with an unauthenticated user over email, they must manually add the email identity to that user record. If the unauthenticated user's email address already exists on a different user record, the agent has the option to merge the two user records.

      Prior to manually adding an email identity or merging two user records, we recommend having your agents perform an identity verification check to prevent social engineering attacks.

    • Use both verified and unverified emails: Email identities are created both for users who are authenticated and have a verified email address in their JWT as well as for unauthenticated users who provide email addresses through bot flows. With this option, agents are more likely to find the end user's email address in the essentials card, and email identities for unverified email addresses are clearly marked as unverified in the Agent Workspace. When needed, agents can ask end users to authenticate if they need to send email follow-ups.

      In the event of a conflict, verified email identities supersede unverified email identities.

  4. (Optional, but not recommended) If you want any user, even unauthenticated users, to be able to claim verified email addresses, select Unauthenticated user can claim verified emails.

    When you select this option, the verification state of email identities collected from messaging channels is no longer trustworthy because an imposter can show up after a user is authenticated and take possession of their email status in a later messaging interaction. This means impersonation attacks are more likely to succeed and agents have limited means of knowing whether the end user is who they claim to be. However, verified email identities still superscede unverified email identities and the email identity is removed from the imposter's user record.

  5. Click Save settings.

Issuing JWTs with email addresses

If you're using email identities, you need to include two additional claims in the JWTs you issue to end users. That means your JWTs will include the following fields:
  • external_id: (Required) This is the unique alphanumeric string that can be used to identify each user. Commonly, these IDs are pulled from external systems. An ID can't exceed 255 characters.

    Zendesk uses the external_id as the primary identifier for users authenticating over messaging. When authenticating a user with a valid JWT, Zendesk first resolves an existing user with the external_id. If no external_id match is found, users are resolved using the email address provided. See Issuing JWTs with email addresses.

  • scope: (Required) The caller's scope of access. The only valid value is user.
  • name: (Optional) The name of the user. Including the name in the JWT payload enables Zendesk to display the user's name in the Agent Workspace and helps your agents provide more customized support.
  • email: (Required) The email address of the user being signed in. Must be unique to the user.

    Set the email address to the user's primary email address in Agent Workspace. The inclusion of secondary email addresses in JWTs isn't supported.

  • email_verified: (Optional) Whether or not the end user in question has proven ownership of the email address. If you want end users to have verified email identities, the JWTs you issue must contain both email and "email_verified": true claims.
For example: 
{
    "external_id": "12345678",
    "email": "janes@soap.com",
    "email_verified": true,
    "name": "Jane Soap",
    "scope": "user"
}

Resolving conflicts between external IDs and emails in JWTs

Zendesk uses the external ID as the primary identifier, with email addresses being used only if no matches are found for the external ID. If, however, an email address presented in a JWT is already associated with a different external ID, Zendesk rejects the JWT and the end user's login attempt fails. When this happens, the conversation begins with the user in an unauthenticated state.

To resolve conflicts between external IDs and emails in JWTs
  • Update the JWT to use a different external_id value or email address.

    OR

  • Delete the user with the conflicting external_id, freeing it up to be used by a different end user. See Delete User in the Sunshine Conversations API.
Powered by Zendesk