Create a Campaign with Campaign Decisioning

Early Access

This is an Early Access feature. To enable it for your account, please contact your MoEngage Customer Success Manager (CSM) or the Support team.

After creating AI Agents to MoEngage, you can include them in messages sent through campaigns and Flows. MoEngage currently supports sending coupons through the following channels:

Channel Name	One-time	Periodic
Mobile Push	Yes	Yes
Web Push	Yes	Yes
Email	Yes	Yes
SMS	Yes	Yes
WhatsApp	Yes	Yes

The process for assigning a campaign to an agent is outlined below. This article guides you through the process of creating campaign decisioning-based campaigns. You can follow a similar approach to assign other supported channel campaigns to an agent.

warning

Warning

Create the Campaign Decisioning agent you intend to use in the campaign before following the steps below. You can use only active Agents in campaigns. For more information, refer to Create an Agent.
On-demand campaigns are not supported for Campaign Decisioning agents. Ensure to schedule your campaign at least 24 hours in advance.
Establish a clear tagging structure before onboarding. Well-defined tags help AI agents categorize content and learn user preferences more quickly, enabling smarter campaign decision-making. For more information, refer to Manage Tags.

Create the Campaign Decisioning

To create the campaign decisioning managed by an AI Agent, perform the following steps:

On the left navigation menu in the MoEngage UI, use one of the following options:
- Click Engage > Campaigns. On the All campaigns page, click + Create campaign. The Create campaign page is displayed.
- Click Create New > Campaign. The Campaign page is displayed.
Select your desired channel (for example, Push) and delivery type (One Time or Periodic).

The process includes three steps:

Step 1: Target Users

Define the basic campaign details and select the AI delivery strategy.

Define Campaign Details: Enter the basic information for your campaign.
Campaign tags: Select or create relevant context labels for the campaign.
Note: You must select at least three tags. These tags provide the features that AI needs to understand the nature of the campaign (for example, Seasonal, High-Value, or Cross-sell) to optimize its decisions effectively.
Configure Delivery Strategy: In the Delivery Strategy section, click the Campaign Decisioning card. The Select agent list will appear.

Note: Selecting Standard lets you manually define your audience segments and delivery rules. For more information, refer to Campaign-Decisioning-Agents.
Select agent: In the drop-down list, click the active AI Agent that will manage the delivery logic for this campaign.
Select Target Audience: Select your audience based on your intended delivery channel.
- Refer to specific channel documentation for detailed instructions on Push, Email, SMS, and WhatsApp setup.
- Note: The audience defined for this campaign must be a subset of the higher-level audience previously defined in your Agent setup.

Step 2: Content

Define the message (in this case, Title, Body, and Images) that the AI will deliver. The Agent evaluates how these messages perform across different user cohorts to refine future deliveries.
You can also add multiple variations of the content, and the most suitable variation will be sent to the user.

For more information on Content for your selected channels, refer to their documentation

Step 3: Schedule and Goals

Finalize the campaign timeline and define the success metrics (KPIs) to track performance.

Schedule:
1. Select the date and time for the campaign to begin. Only future dates are supported to allow the AI to prepare initial predictions.
2. Optionally, select Best Time to Send to automatically schedule delivery when each user is most likely to engage. This ensures your entire campaign, from audience, channel, content, to schedule, is fully optimized for performance.
Conversion Goals: Add the goals you want to track. These serve as the Rewards the AI Agent uses to learn which users respond best to this campaign.

FAQ

arrow_drop_down Why must campaigns be scheduled 24 hours in advance?

The Decisioning Agent requires this window to analyze complex user data and campaign features. To ensure the AI can accurately identify the best matches for every user, campaign settings are locked at midnight to allow for algorithmic processing.

arrow_drop_down Can I edit today's campaign content right before it sends?

While you can make changes, the Decisioning Agent will not recognize them immediately. The AI optimizes delivery based on a synchronization that occurs at midnight; therefore, you must apply content updates before the daily processing begins.

arrow_drop_down Do AI Decisioning campaigns support A/B testing, and how does it function?

Yes. AI Decisioning uses Contextual Multi-Armed Bandits (CMAB), a more sophisticated approach than the standard "vanilla" bandits used in traditional campaigns. While standard testing looks for a single "winner," our Agent evaluates each user's unique context to assign the most relevant variation. It first identifies the optimal campaign for the individual and then selects the specific content variation most likely to maximize the agent's reward.

arrow_drop_down Does AI Decisioning support Locales, and how does it work?

Yes. Locale functionality in AI Decisioning campaigns operates identically to Standard campaigns. Locales are automatically assigned based on the user property rules defined during the locale setup.

arrow_drop_down What is the difference between Agent and Campaign segments?

The Agent Segment is the broad audience the AI learns from to maximize long-term goals. The Campaign Segment provides tactical guardrails to ensure message relevancy. For example, while the Agent analyzes all users, you might restrict a "Summer Dresses" campaign specifically to women. These campaign-level rules act as hard boundaries that the AI cannot cross.