Adobe Analytics
Run reports and top-item reports, read report suites, segments and metrics, and create segments.
An Adobe Analytics connection comes in two forms. Adobe Analytics gives a playbook's functions the Analytics 2.0 API, from a document Auto ships of twelve operations, under your company's global company ID; it connects with a Server-to-Server credential of an Adobe Developer Console project, or by signing in with your Adobe ID. Adobe Analytics MCP gives them Adobe's Analytics MCP server, which answers questions about report suites, dimensions, metrics and reports with the access your product profile gives; you sign in with your Adobe ID, and that connection lasts as long as its sign-in, since Adobe's server issues no refresh token.
Available on Auto Cloud and on self-hosted runtimes.
What a playbook can do with Adobe Analytics
- Run a report and a top-item report over a report suite's dimensions and metrics for a date range.
- Read report suites, dimensions, metrics, segments, calculated metrics, date ranges and projects.
- Create a segment.
- Through the MCP server: ask about report suites, dimensions, metrics and reports.
The report operation is marked as a tool that only reads, so a lookup function can run it, though Adobe's API takes it as a POST.
The tools of Adobe Analytics, as the connection lists them
Tools that read: findCalculatedMetrics, getDateRanges_1, dimensions_getDimensions, getMetrics, projects_getProjects, runReport, runTopItemReport, segments_getSegments, segments_getSegment, getCurrentUser and getSuites_1.
Tools that act: segments_createSegment.
Adobe Analytics MCP: its MCP server lists its own tools when a function runs, and once you connect, your agent lists them with list_connection_tools.
An example workflow
A daily check of a report suite's key metrics against what your analysts expect. A schedule trigger runs the workflow at 07:00. A lookup function runs one report of yesterday's visits, orders and revenue by marketing channel. A computation function compares each figure with the same weekday of the last four weeks and flags the ones more than a set share away. A reasoning function reads the flags with the notes in its prompt, the campaigns that launched, the tracking changes the team made, and writes which deviations have a known cause and which need a look. An interaction function posts the note to the analytics channel through a Slack or Microsoft Teams connection. A prediction function can forecast the week's revenue from the figures the playbook has recorded, with a likely range. Every report's rows are in the run's history in Studio.
How to connect Adobe Analytics
Adobe Analytics
- Give the client id and secret of a Server-to-Server credential of an Adobe Developer Console project with the Adobe Analytics API.
- Sign in with your Adobe ID and allow Adobe Analytics. The consent screen shows the permissions listed below.
Adobe Analytics MCP
- Sign in with your Adobe ID and allow the Analytics MCP server. The consent screen shows the permissions listed below.
Then:
- Ask your agent to connect Adobe Analytics to Auto, or open Settings, Agents, Connections in Auto Studio and choose Adobe Analytics or Adobe Analytics MCP.
- On the Studio page, check the service and its access, choose which playbooks may use the connection, and approve it. Give a key or a client secret there, never in a conversation with your agent.
- Have your agent check one small read through the connection before a function depends on it.
Adobe Analytics asks for one value when you connect:
| Value | What it is |
|---|---|
global_company_id | Your global company ID, which https://analytics.adobe.io/discovery/me answers. |
What it needs
- An Adobe Analytics company and a product profile that gives your account, or the technical account of the credential, the report suites the playbook reads.
- For the credential: a project in the Adobe Developer Console with the Adobe Analytics API and a Server-to-Server credential, whose client ID and secret you copy.
- Your agent connected to Auto and signed in as you, with permission to change your organization, if your agent adds the connection.
- On a self-hosted runtime, for the sign-in to the 2.0 API, an OAuth client registered at Adobe IMS for the runtime; the MCP server lets Auto register a client itself.
Permissions Auto asks for
Adobe Analytics, with client credentials:
- Read the unique identifier of the Adobe account the connection signs in as.
- Keep a session at Adobe IMS for the technical account of the credential.
- Act as the Adobe ID of the account in calls to Adobe APIs.
- Read the Adobe organizations the account belongs to.
- Read the Adobe product profiles and Adobe Analytics permissions of the account.
Adobe Analytics, by sign-in:
- Read the unique identifier of the Adobe account the connection signs in as.
- Act as the Adobe ID of the account in calls to Adobe APIs.
- Read the Adobe organizations the account belongs to.
- Read the Adobe product profiles and Adobe Analytics permissions of the account.
- Read the job function in the Adobe profile of the account.
- Keep access to Adobe Analytics when nobody is signed in.
Adobe Analytics MCP:
- Read the unique identifier of the Adobe account the connection signs in as.
- Act as the Adobe ID of the account in calls to Adobe APIs.
- Read the Adobe product profiles and Adobe Analytics permissions of the account.
Good to know
The MCP connection ends when its sign-in does, since Adobe's server issues no refresh token; it suits work your agent does with you, where the 2.0 API connection with a Server-to-Server credential suits a scheduled workflow. A report's rows are kept in the run's history, where people with access to the playbook can read them.