Why AI Decisioning Engine?
The AI Decisioning Engine (formerly Inspire by Building Blocks) is designed to enhance marketing effectiveness through personalization, automating labor-intensive tasks for marketers. By demystifying AI, it provides transparency and control over AI-driven operations, making your marketing strategies not only efficient but also easily adjustable to align with your business goals. This ensures that your marketing efforts are both effective and insightful, empowering you to make informed decisions backed by data.
Understanding the AI Decisioning Engine
The AI Decisioning Engine operates as an advanced analytical layer atop the Customer Data Platform (CDP), with the primary function of enriching customer profiles. By examining web tracking events like product detail page (PDP) views, add-to-cart actions, and purchase histories, it transforms raw data into actionable insights that enrich the profiles within the CDP. This enrichment is critical as it provides a nuanced understanding of each customer's behavior and preferences.
These enriched profiles serve as the foundation for personalized customer interactions. They empower marketers to tailor touchpoints across various channels, enhancing engagement through personalization tools available within the Mobile Marketing Cloud (MMC). This includes personalized (email) campaigns and web personalizations, allowing businesses to deliver relevant content that resonates with their audience effectively.
Cockpit: Empowering Control Over AI Personalization
At the core of our competitive edge is the Cockpit, a versatile tool that provides an unparalleled level of control over AI personalization. This is a unique feature not commonly seen in competitive tooling. The Cockpit allows businesses to be in the driver's seat of their personalization strategy, offering comprehensive control and flexibility over the AI engine's operations.
-
Personalization and Recommendation Creation: Businesses can design new personalizations or recommenders that align with specific objectives, whether for a particular campaign or touchpoint.
-
Detailed Control Options: Users can tailor how products are recommended, down to the user-level specificity, allowing for strategic product positioning based on business priorities, such as prioritizing high-margin in-house brands or focusing on particular product types. Control extends to excluding previously purchased items, ensuring every interaction remains relevant.
-
Quality Assurance Tools: The Cockpit provides statistical tracking and insights for quality control. Businesses can view data on frequently recommended products and how recommendations distribute across attributes like brand or category, allowing for strategic refinement of AI models to enhance personalization accuracy.
Key personalizations of the AI Decisioning Engine include:
-
AI-Driven Recommendations: The engine specializes in generating precise, personalized product suggestions that enhance customer engagement. From homepage suggestions and newsletter recommendations to PDP complements like "complete your order" and tailored alternatives, these recommendations adjust dynamically based on detailed customer profiles and actions.
-
AI Predictive Insights: Beyond recommendations, the engine forecasts critical purchasing behaviors. It anticipates product preferences and optimal next purchase moments. These AI-driven predictions empower marketers to seamlessly orchestrate next best actions, maintaining customer engagement and guiding them through their buying journey.
-
Behavioral Triggers: These functions automatically detect actions like abandoned carts, abandoned checkouts, and abandoned browsing sessions — including price drop detection — initiating targeted responses within the CDP. They support immediate personalization strategies for follow-up campaigns and web interactions, available in both batch and near-real-time processing.
-
Next-Generation AI Engine: Powered by a neural network transformer model, the next-generation AI engine takes personalization further by learning from a broader range of customer touchpoints. Launching in June 2026, it enables more accurate product recommendations, with future capabilities extending to timing predictions and churn prevention.
Packaging
Basic behavioral triggers, RFM analysis, item enrichments, attribute tagging, frequently bought products, attribution, and AB testing are accessible with the MMC PRO package. More advanced AIDE personalizations such as personal recommendations, contextual product recommendations, event-based retargeting, next purchase moment predictions, near-real-time processing, and the next-generation AI engine are available through the MMC Enterprise package.
The MMC Advanced tier does not include AIDE personalizations.
Required Applications
To fully leverage AIDE, ensure integration with CDP and Webtracking, facilitating events such as PDP View, Add to Cart, Remove from Cart, and Purchase lines using the Retail & Ecommerce adapter. A product feed is also essential.
Feature Overview
-
Cockpit - This feature empowers users to oversee their personalizations and establish guidelines within which the AIDE operates, allowing businesses to strategically apply AI while maintaining control over quality. Cockpit enables customers to be in the driver's seat, giving them full control over the AI engine and tools for quality assurance. Specific capabilities include:
-
Personalization and Recommendation Creation: Users can create new personalizations or recommenders tailored to specific goals, such as a recommender for a specific campaign or touchpoint.
-
Detailed Control Options: Users can configure which products can be recommended, even down to a user level, allowing for product prioritization based on business goals. For example, they can boost certain products, such as in-house brands with higher margins, or highlight specific product types. Controls also include excluding previously purchased or recommended products to avoid redundancy and ensure relevance.
-
Quality Assurance Tools: Cockpit offers statistical tracking to support quality control, allowing users to see data like the most frequently recommended products and how recommendations distribute across specific item attributes (e.g., brand, category).
-
Behavioral Insights: Users can inspect individual profile behaviors and the resulting AI-driven personalizations, offering transparency into the AI's decision-making process.
Insights gained from these quality assurance tools can be used to refine the AI model, enhancing personalization accuracy through strategic adjustments.
-
Available Personalizations
-
Behavioral Triggers, RFM, Item Enrichment, Attribute Tagging & Frequently Bought Product
-
Behavior Triggers (Available for MMC PRO Customers): Detects abandoned carts, abandoned checkouts, and abandoned browsing sessions, triggering actions within the CDP for use in email campaigns and web personalizations. Price drop detection is available for abandoned cart and abandoned checkout triggers, enabling re-engagement when items in a customer's cart or checkout drop in price. Near-real-time processing is available for MMC Enterprise customers, enabling immediate trigger detection and response.
-
RFM Analysis (Available for MMC PRO Customers): Conducted and stored at the profile level, facilitating customer segmentation based on value and recency.
-
Item Enrichment (Available for MMC PRO Customers): Automatically enriches product feed items with recent purchase and popularity data, useful for identifying best-sellers in product segments.
-
Attribute Tagger (Available for MMC PRO Customers): Helps identify customer product preferences for targeted marketing by tagging users based on their item attribute engagement patterns.
-
Frequently Bought Product (Available for MMC PRO Customers): Identifies frequently bought products for each user based on purchase history, enabling personalized reorder suggestions and replenishment campaigns.
-
-
Analytics & Testing
-
Attribution (Available for MMC PRO Customers): Attributes conversions and engagement to marketing touchpoints, providing insight into which interactions drive results across the customer journey.
-
AB Testing (Available for MMC PRO Customers): Enables experiment management with variant assignment, metrics tracking, and performance evaluation to measure the impact of personalization strategies.
-
-
Recommendation Personalizations
-
Personal Recommendations (Available for MMC Enterprise Customers): Tailor content for users based on their behavior.
-
Example Use Cases: Homepage recommendations, Newsletter campaign recommendations.
-
-
Contextual Product Recommendations (Available for MMC Enterprise Customers): Offers personalized suggestions based on context.
-
Example Use Cases: PDP Complements ("Complete your order" / "Others also bought"), PDP Alternatives ("Maybe this is more to your taste").
-
-
Event-based Retargeting (Available for MMC Enterprise Customers): Personalized suggestions based on customer triggers.
-
Example Use Cases: Abandoned cart/browse/checkout recommendations, After purchase cross-sell, Canceled purchase follow-up, Purchase intent recommendations.
-
-
-
Predictions
-
Next Purchase Moment (Available for MMC Enterprise Customers): Predicts optimal re-engagement times for efficient warming up of potential repurchases.
-
-
Next-Generation AI Engine
-
Transformer Recommender (Available for MMC Enterprise Customers, launching June 2026): A neural network-based recommendation engine that learns from a broader range of customer touchpoints for more accurate and nuanced product recommendations. Future capabilities will extend beyond product recommendations to include timing predictions and churn prevention, providing a unified AI engine for the full customer lifecycle.
-
By integrating AIDE into your marketing strategy, you not only personalize customer interactions effectively but also harness the full power of your data, facilitating impactful and dynamic customer journeys.