Campaign Use Case Examples

This page provides real world examples of Fabric Campaigns.

Campaign Component Descriptions	Example Use Case A: Healthcare	Example Use Case B: Healthcare
Campaigns address multi-layered business problems.	A healthcare provider wants to increase member satisfaction and savings by promoting preventive opportunities.	A healthcare provider wants to increase membership enrollment to medical plans.
Cohorts are defined as the different groups that are targeted by Campaign Mission Goals.	One Cohort is defined as healthcare members who are covered by Medicare.	One Cohort is defined as individuals who access the provider’s website but are not currently enrolled. Another Cohort is defined as individuals who were formerly but are not currently enrolled.
Goals with measurable outcomes (KPIs) are defined to quantify the success of the Mission for the targeted Cohort groups.	One Goal is defined to increase the number of Medicare-covered members who receive the annual flu shot by 50%.	The Goal (KPI) for the first Cohort is to increase the conversion rate for website visitors by 50%. The Goal (KPI) for the second Cohort is to have 30% of former members re-enroll.
Missions are designed to address a specific Goal for a specific Cohort using a number of strategies or Interventions.	The Mission is to get 50% more Medicare-covered members to get flu shots using the following Interventions: - nurse phone call reminders - text message reminders - email reminders.	A Mission for each Goal-Cohort is configured. Interventions for the first Goal-Cohort might include: - To present 3 different landing pages with different designs and messaging - To use different outreach channels (Instagram, Facebook, TikTok)
Each Interventions is configured for: Cost, Timeout, Pre-condition, Wait Time, Effects, and Action	Cost, Wait Time, and Effect are required to run a Simulation. Pre-condition and Effects are defined using low-code expressions.	Cost, Wait Time, Pre-condition, and Effects are configured. Simulations are run.
A Mission Simulation is run using synthetic data. The simulation randomly assigns Interventions to make predictions about the possible outcomes for cohort members with shared attributes.	The simulation randomly assigns Interventions to make predictions about the possible outcomes for the synthetic Cohort members generated by Fabric.	The simulations randomly assign Interventions to make predictions about the possible outcomes for the synthetic Cohort members generated by Fabric.
Simulation results provide ranked insights about the success of Interventions on members of the Cohort who share attributes. Domain experts refine Intervention Plans.	Grouping the Simulation outcomes by age group and gender, the domain expert views how each Intervention performed. They re-rank and remove Interventions based on performance and expertise.	The domain experts look at the Intervention plan rankings based on a variety of attributes like income, age, and type of employment.
Interventions are reconfigured and simulations are run until the optimal Interventions are determined.	Because the text messaging Intervention achieved sub-optimal results, that Intervention is removed from the Mission.	The domain experts re-rank the Plans and make modifications to the Mission configuration to devise the optimal Intervention strategies.
The Intervention Actions are developed and packaged as Skills (Models) in Fabric and selected prior to Campaign or Mission deployment.	An AI model that is developed to send emails that contain a button that takes them directly to a flu shot registration page to healthcare members. The model is packaged into a Fabric Skill and assigned to the Intervention.	An AI Skill and Model are packaged for each Intervention. The developer returns to the Intervention configuration and selects a Skill to serve as the “Action” for each of the Mission Interventions.
The Mission is deployed independently or with other Missions in a Campaign deployment.	Subsets of the Cohort are assigned Interventions randomly at first. Feedback is collected and converted into Rewards to train the Model. As the Model learns from the Feedback/Rewards, the interventions are targeted to members who are likely to respond positively based on similar characteristics.	For each runtime cycle the Model becomes more adept at assigning interventions to Cohort members who respond positively to the intervention. When all of the Cohort members have been served an Intervention, the Campaign or Mission run is completed.
The Mission or Campaign returns the KPI measures as Profiles are updated.	The business users assess the success of the AI driven solution to their business problem by determining if the KPIs were met.	The business users assess the success of the AI driven solution to their business problem by determining if the KPIs were met.