AI decision shape
Overview
The AI decision shape allows you to split the journey into different paths by letting AI analyze contact data and determine which criteria apply to them.
EX: AI can assess a customer’s activity and field data to predict their likelihood of canceling their account.
IMPORTANT: This feature is currently available for free until May 24th 2025.
After this date, it will be offered as an add-on. To continue using this feature after this specific date you will be able to add this to your current plan under your Plan & Billing page.
Accessing the AI decision shape
The AI decision shape is available in all journeys. To add one, edit the journey > click the + icon to add a new shape > AI decision.
TIP: To help make Ortto AI smarter and more reliable across the app, we strongly recommend adding descriptions to fields and activities.
Learn more about field and activity descriptions.
Set up the AI decision shape
After adding the AI decision shape to a journey, the next step is to configure it. The AI decision shape consists of three main sections that you can customize:
Prompt
In the Prompt input, clearly describe what you want Ortto AI to decide. Different outcomes will be determined based on this description.
Context
Based on your prompt, our AI will automatically generate relevant context, split into different groups that you can customize. These groups should contain the information from your account that the AI will consider when making decisions.
Examples of these groups include:
- Customer engagement
- Customer satisfaction
- Customer support
- Customer demographics
These are just examples, AI-generated groups will vary based on your description.
You can add more relevant groups to this section. To do this, click + Add group > enter a name > select the fields and activities relevant to the group.
TIP: The more context you provide to the AI, the better the results. There is a limit of 20 fields and activities, so we encourage you to include as many relevant ones as possible to improve the accuracy of the AI analysis.
Outcomes
Based on your prompt, our AI will automatically generate potential outcomes. Each outcome represents a possible path for each contact to follow based on the AI’s analysis.
We require at least 2 outcomes, and we support a maximum of 20 outcomes.
You can manually add more outcomes by clicking + Add outcome. Then, enter a Title and the Criteria for that outcome.
Once everything is set up, click Finish setup to add the shape to your journey.
AI decision activity
Each time a contact passes through an AI decision shape, an AI decision activity is automatically recorded with three key attributes:
- Campaign name: The journey's name.
- Outcome: The AI-determined outcome.
- Reasoning: The AI's analysis behind the decision.
You can view this activity in the contact profile, the journey's Activity tab, and use it in filters and reports like any other activity.
Didn't match outcome
Contacts will follow the Didn't match path when the AI cannot confidently make a decision based on the available data. This may happen due to insufficient data in the contact profile.
EX: If the contact hasn't performed any of the activities selected in the Context section or if the data is too unclear for the AI to process your prompt.
If you believe the AI should be making a decision for a specific contact but isn't, contact our Support team, and we will investigate further.
Use cases
The AI decision shape helps optimize your customer journeys by segmenting contacts based on key behaviors. Here are some common use cases:
Customer cancellation risk
The AI decision shape identifies contacts at risk of canceling their service.
- Low risk: Highly satisfied customers, unlikely to cancel.
- Medium risk: Customers showing signs of dissatisfaction but still engaged.
- High risk: Customers likely to cancel, allowing proactive actions like discounts or feedback requests.
Purchase intent
The AI predicts the likelihood of a customer making a purchase based on their interactions.
- High intent: Customers likely to buy soon based on engagement.
- Low intent: Customers with limited interest or interaction.
Subscription renewal probability
The AI evaluates the chances of a customer renewing their subscription.
- Likely to renew: Customers expected to renew, perfect for reminder emails.
- Unlikely to renew: Customers who may need additional incentives or offers.
Lead scoring
The AI categorizes leads based on their engagement and fit, guiding you on how to approach them.
- High-quality lead: Highly engaged leads, ready for sales outreach.
- Low-quality lead: Leads requiring further nurturing before sales engagement.