Author and share your own playbooks (Beta)
The Playbook Authoring Framework, powered by AI Assistant in Adobe Experience Platform, allows you to create, manage, and share playbooks efficiently within Adobe Experience Platform.
The framework follows a three-step process:
-
Metadata capture: Use AI Assistant or the webform to capture playbook metadata.
-
Technical association: Add specific technical assets such as journeys or audiences to the playbook. You retain full control over the playbook creation process within your development sandbox, ensuring alignment with your schemas and other unique data structures.
-
Playbook distribution: Share playbooks across different organizations. For example, ACME’s Martech Center of Excellence in Germany can create a “golden” playbook and distribute it to regional organizations in Thailand, Australia, etc. to help standardize the marketing use case.
Create a Playbook
You can create a playbook in two ways: either using the AI Assistant or manually. Read the following sections to learn how to create a playbook with AI Assistant.
Playbook overview
Follow these steps to create a playbook with the AI Assistant:
In the left navigation panel, select Playbooks.
Select New Playbook, and then select Generate playbook with AI Assistant.
Use the prompt field to describe the use case. For example:
“Engage ACME customers who browsed running shoes but did not complete the purchase.”
Select Generate to create the playbook metadata.
Once generated, select Edit to modify the generated title, description, and metadata as needed.
To ensure that the data engineers have all the necessary details to set up the use case, fill out the Playbook detail section. While optional, these fields help capture key information, making it easier to connect the right technical components. Select Edit to add values to the following fields:
- Industry
- Target audience
- Marketing channel
Once the metadata is generated, select Edit journey map to adjust the steps in the journey map as required.
Then, proceed to associate the playbook with technical assets. To create a playbook manually, select Create playbook manually.
A blank playbook template appears. Fill out details such as Title and Description. You can also edit the journey map to add events and touchpoints as needed.
Associate playbook with technical assets
Regardless of whether you create a playbook manually or with the AI Assistant, you must associate it with the required technical assets. Navigate to the Technical Assets tab and select the required product. For this example, select Journey Optimizer.
Choose Select an Asset to associate this playbook with a journey as shown in the image below. Then select Publish playbook to finalize the playbook.
Once published, the playbook automatically extracts and associates the journey’s schema and audience details.
All the created playbooks are available in the Your Playbooks tab.
You can select any playbook from the catalog to create instances for reuse. Refer to the documentation to learn how to create instances.
Example prompts
The AI Assistant can process various prompt structures and extract key details while filtering out unnecessary information. Below are some examples of user prompts and how the system interprets them.
Example 1:
“Create a campaign titled “Complete the Look” in order to increase sales and CLV. The campaign encourages customers purchased kitchenware or furniture to complete a complementary purchase via personalized recommendations and offers related to their purchase. First message the customers with product recommendations. If they don’t make any purchases within 7 days, they receive a second message with product recommendations and offers. Use push notifications and email to contact the customers. Target customers who made a purchase in last 7 days in kitchenware or furniture category and have not been targeted in last 30 days. As part of the campaign, we want to measure KPIs such as clicks (email, app, sms, push), CTR, E-Wallet CTR, AOV Conversion.CLV Revenue, Total Purchase events (in-store, digital, call center).”
Example 2:
“Project Name: Fashion Newsletter
Background: (proactive or solving for problem): A journey designed to send fashion newsletters to ACME customers who have subscribed for newsletter communication.
Objective: Send fashion newsletter emails to ACME customers who subscribed for communication.
Promotional details: Customer receives fashion news in the email channel weekly. The email should be personalized and content variations based on gender, language and market.
Project channels/Touchpoints: Email
Target audience: Customers who have subscribed to ACME fashion newsletter communications.
Target KPIs/Engagement Metrics/ROI: 1. Increase Revenue from Products. 2. Drive Customer loyalty.”
Example 3:
“Nudge shoppers to buy products during an ongoing product promotional campaign.
Engage with shoppers during an ongoing promotion by sending appropriate communication through email, SMS, or push notifications to buy products. Send them a reminder email after 24 hours of them not engaging with the promotion.”
Example 4:
“Sell shoes to high school players.”
The AI Assistant removes all unnecessary details such as “Project Name” or “Background.” It extracts the key elements such as “target audience”, “campaign goal”, and “marketing channel” and works with any input style.
These examples demonstrate how AI can refine and extract essential details from user prompts.
Content guidelines and moderation
When creating playbooks, be mindful of the language and content you include. Playbooks are visible across your organization, and any offensive or inappropriate content that users flag.
Flagging and review process
If a playbook contains inappropriate or offensive content, Adobe automatically receives a report for review. Adobe reviews the flagged content, notifies the customer if it is deemed inappropriate, and removes the playbook.
Share playbooks across sandboxes share-playbooks-sandboxes
When you create and publish a playbook in one sandbox, it automatically becomes available across all sandboxes within your organization. This feature eliminates the need for manual sharing and allows you to create instances of the playbook in any other sandbox seamlessly.
If any fields are missing from your union schema, a dialog box highlights them during the import.
Sharing your playbooks across organizations sharing-playbooks-organizations
Sharing playbooks across organizations helps ensure consistency and efficiency when multiple teams need to follow the same best practices. To share a playbook from one organization to another, follow these steps:
-
Log into the source organization: Navigate to the organization that contains the playbook you created and want to share from the Your playbooks tab.
-
Publish the playbook: If the playbook isn’t already published, you must publish it before sharing.
note note NOTE A partnership must be established between the source and target organizations to enable playbook sharing. Learn how to create an organization partnership request. -
Initiate the share: Once the playbook is published and a partnership is established, select Share Playbook.
-
Select the target organization: Choose the organization you want to share the playbook with when prompted.
-
Confirm and share: Confirm your selection. You will receive confirmation messages indicating successful sharing.
-
Verify the target organization: Log into the target organization to verify that the playbook is available.
-
Import the playbook: Select Import to bring the playbook into the target organization. You can view it in the Playbooks tab.
If the playbook doesn’t appear, ensure it is published and that the organization partnernship is active.
Required permissions required-permissions
To access the sandbox and use this feature, you need the following permissions:
Sandbox permissions
These permissions are required to access the sandbox environment where the feature exists:
-
Manage sandbox
-
View sandbox
-
Package sharing permissions:
These permissions are required for internal sharing functionality:
Use these permissions to:
- Enter the sandbox environment
- Access the feature within the sandbox
- Manage and share packages as needed
These permissions are located in the Sandboxes section of the permissions list.
Journeys and related objects - permissions
When building Journeys that use Playbooks, you might reference other objects, such as Channels, Audiences, and other entities. Each of these objects has its own permission set.
These are the key permissions for Journey-related actions, such as:
- View journey
- Manage journey
- Permissions related to objects like Audiences, and Channels
You also need the following audience permissions:
- Segment read
- Profile read
- Dataset read
As Journeys are highly flexible and can involve many interconnected objects, their full permissions are documented separately and can vary based on your particular use case.
Next steps
Now that you understand how to create, publish, and share playbooks using the AI Assistant, learn how to get started with the available playbooks and choose the right one for your use case from Playbooks List.