Growth Play Agents are AI-powered assistants embedded directly in the Growth Engine. They provide contextual guidance as you discover, configure, and optimize Growth Plays in BlueConic, reducing the need for manual setup and helping you act faster on your goals.
Five agents are available, each designed for a specific stage of the Growth Play workflow:
Agent | Location | Purpose |
Growth Play LaunchPad | Recommends Growth Plays based on your goals | |
AI Canvas | Generates audience segments from natural language prompts | |
Use the Listener Agent | Guides configuration and troubleshooting within an active Growth Play | |
Use the Listener Agent | Sets up listener configurations from a plain-language description | |
AI Canvas | Creates timeline event rollups and profile properties from natural-language descriptions | |
AI Canvas | Sets up connections (Meta Ads, Klaviyo, Salesforce Marketing Cloud, and more) through guided, conversational setup |
Before you begin
For an overview of Growth Plays, see Growth Plays overview.
Use the Advisor Agent
The Advisor Agent is available in the Growth Play LaunchPad. Use it to identify the Growth Plays most relevant to your business goals and current BlueConic setup.
When you describe what you want to accomplish, the Advisor Agent analyzes your intent and surfaces recommended Growth Plays directly in the LaunchPad.
How to use the Advisor Agent
From the LaunchPad, enter a business goal or use case into the prompt field.
Examples include:
Recover abandoned carts
Improve customer retention
Increase loyalty engagement
Find high-intent shoppers
Reduce customer churn
Based on your request, the Advisor Agent automatically:
Recommends relevant Growth Plays
Highlights suggested plays within the LaunchPad
Explains why a Growth Play is a strong fit
Provides a confidence or match score for the recommendation
Note: For example, when asking about cart recovery opportunities, the Advisor Agent may highlight the Cart Recovery Growth Play and explain why it aligns with your objective.
Use the Segment Agent
The Segment Agent helps automate audience creation within the AI Canvas.
Instead of manually building audience logic from scratch, you can describe the audience you want to target and the Segment Agent generates recommended segments automatically.
How to use the Segment Agent
In the AI Canvas, enter a natural language description of the audience or goal you want to target. Examples include:
Create a segment of high-intent shoppers
Find customers likely to churn
Build an audience for cart recovery
Identify loyalty members close to the next tier
Create a segment for recent purchasers
The Segment Agent automatically recommends audience criteria based on your description.
Refining segments
After a segment is generated, you can:
Modify the segment manually
Continue refining it using natural language
Adjust rules and attributes directly in the canvas
Combine AI-generated logic with traditional controls
This allows you to move quickly while still maintaining full control over your audience configuration.
Use the Coaching Agent
The Coaching Agent operates directly inside the AI Canvas and provides contextual guidance throughout Growth Play configuration.
Unlike the Advisor Agent, which focuses on recommending Growth Plays, the Coaching Agent helps you configure and complete the selected play.
How to use the Coaching Agent
The Coaching Agent is triggered through natural language prompts entered within the AI Canvas.
Once prompted, the Coaching Agent analyzes the current state of your canvas and provides recommendations and guidance directly in the side panel.
The Coaching Agent can:
Recommend next configuration steps
Explain objects and settings
Identify incomplete setup steps
Help troubleshoot issues
Guide you through implementation tasks
Suggest actions based on your current progress
For example, after confirming required integrations or data sources, the Coaching Agent may recommend configuring audience segments as the next step in the workflow.
Interact with the Coaching Agent
You can interact with the Coaching Agent using natural language prompts.
Examples include:
What should I configure next?
Why is this step incomplete?
Explain this object
Help me finish this setup
How do I improve this audience?
Responses appear in the AI Canvas side panel and are contextual to your current Growth Play configuration.
Use the Listener Agent
The Listener Agent helps you configure listeners in AI Canvas without needing to work through each technical setting yourself — which signals to track, how quickly older data should fade, which values to include or exclude, and so on.
How to use the Listener Agent
In the AI Canvas, describe what you want the listener to capture, in your own words. Examples include:
Rank interests based on pages visited and content categories
Detect when a visitor's mouse moves toward the browser's exit area
Flag a visitor as inactive after a period with no clicks, scrolling, or form input
Track when a visitor abandons a multi-step form partway through
The Listener Agent proposes a complete, ready-to-use listener configuration directly on the canvas, along with a plain-language explanation of why it chose those settings. New listeners are created in an off state. The agent will ask you to confirm turning it on before it starts capturing data
Review and confirm a configuration
Nothing is saved until you approve it. Once the agent proposes a configuration, you can:
Review the settings and the agent's explanation for each choice
Ask for changes in your own words
Confirm the configuration to apply it
This keeps you in control of the final setup while removing the manual work of tuning individual listener settings.
Supported listeners
The Listener Agent currently supports:
Rule Based Form listener
Funnel Step listener
Event Trigger listener
Visual Form listener
Scoring listener
Recently Viewed Items listener
Behavior listener
Campaign Tracker listener
Geolocation listener
Engagement listener
Keyword Interest Ranking listener
Product Recommendation Events listener
Logout Detection listener
AdBlock Detection listener
Privacy: Consents and Refusals listener
European Visitors: Only Profile After Consent listener
Creating a profile property automatically
When you ask the Segment Agent to build a segment that needs a profile property that doesn't exist yet, it no longer stops there. It proposes the missing property — including a name, type, and data sensitivity — and lets you adjust the proposal or point it at an existing property instead. Nothing is created until you confirm. Once you do, the property is created and the segment continues building using it.
Use the Rollup Agent
The Rollup Agent helps you turn timeline event data into a reusable profile property, directly from AI Canvas, without setting up the timeline event rollup by hand. Describe the aggregated data you need — the event type, any conditions, and how to aggregate it — and the Rollup Agent creates the profile property, the rollup that fills it, and, when you're building a segment, the segment criteria that use it.
How to use the Rollup Agent
In any AI Canvas chat, describe the aggregated data you need. Examples include:
Create a segment of customers who spent more than $100 in the past two weeks, based on Order timeline events
Add a property that totals a customer's order count over the last 30 days
Note: The Rollup Agent works with what you already have configured:
It can't create a new timeline event type — make sure the one you want to aggregate is already being tracked before you ask for a rollup.
It checks your existing rollups first and reuses or updates one where it can, rather than creating a duplicate.
Review and confirm a rollup
Nothing is saved until you approve it.
The Rollup Agent works in Basic mode: event type, conditions, and mappings.
Each rollup it creates is placed on the canvas in the Unify column and stays inactive until you turn it on, so you can open it in the standard rollup editor and review the full configuration first.
If your aggregation needs SQL, the agent tells you and points you to the Timeline event rollups screen instead — it doesn't edit SQL-mode rollups, so it never overwrites an existing SQL definition.
If you don't have the permission needed to create the rollup or property, the agent still shows you the full proposal and names the exact permission you need, so you can request it or ask a colleague to finish the configuration.
Use the Connections Agent
The Connections Agent sets up a connection to an external platform directly from AI Canvas, instead of walking through the connection setup screens by hand. Describe what you want to connect, and the agent asks what it needs, handles the platform login when one's required, and keeps any private credentials out of its own context.
Supported connections
The Connections Agent currently supports:
Amazon Ads
ActiveCampaign
Airship
Attentive
BigQuery
Braze
Cart Abandonment Tracker
Criteo
Databricks
Dynamics CRM
Iterable, Klaviyo
LinkedIn
Magento
Mailchimp
Marketo
Microsoft Advertising
Product Collector
Redshift
Salesforce
Salesforce Marketing Cloud
Shopify
Snowflake
The Trade Desk
TikTok Ads,
VTEX
Webhook (HubSpot)
How to use the Connections Agent
In any AI Canvas chat, describe the connection you want. Examples:
Connect my Klaviyo account so I can add audiences to this Growth Play
Set up a webhook connection to bring in event data from our checkout flow
The agent asks for what it needs one step at a time, using dropdowns, inputs, and connection buttons instead of expecting you to type every value freeform. When a platform requires you to sign in, the agent redirects you to that platform's own login page.
Note: Entering credentials securely
Any private information — an API key, secret, or token — opens in its own secure entry field, separate from the chat. Values you enter there go straight into the connection's configuration; the agent never sees or stores them in the conversation.
Mapping data
If the connection needs a mapping step, the agent surfaces it as its own screen once the rest of the setup is complete, so you review and confirm the mapping before the connection goes live.This is an additional way to set up these connections — the existing connection setup screens are still there if you prefer them.
Next steps
After working with Growth Play Agents, consider the following:
Review the segments generated by the Segment Agent and determine whether additional refinement is needed before activation.
Evaluate the confidence scores provided by the Advisor Agent to prioritize which Growth Plays to run first.
Review the configuration proposed by the Listener Agent before confirming, especially for listeners that affect data capture broadly (e.g., Privacy or Consent-related listeners).