Enabling CX AI Integration

In the era of mass adoption of AI technology, SAP CPQ leverages artificial intelligence to enhance the customer experience. CX AI (Customer Experience Artificial Intelligence) uses machine learning to analyze data and provide recommendations for cross-sell products and discounts on quotes based on a customer's past purchases. CX AI's pricing recommendation feature analyzes historical transaction data to provide optimal pricing recommendations. This can help businesses drive more sales, increase margins and improve customer satisfaction.

To leverage AI-powered recommendations in CPQ, the system must be integrated with SAP CX AI, a cross-platform application. The integration pulls data from SAP CPQ, applies its AI models to analyze this data, and provides recommendations back to SAP CPQ. This data may be related to past sales, customer behavior, and product configurations.

CX AI integration works with Quote 1.0 and Quote 2.0 engines.

All CX AI integration settings are in the Providers section.

In SAP CPQ, ingestion mapping is a process that is part of the integration with SAP CX AI. This process involves mapping data from SAP CPQ to the relevant fields in SAP CX AI. This data is transferred between the two systems to train their respective models.

To ensure reliable model training, mark all quote statuses you wish to send to CX AI. Make sure all selected statuses are final quote statuses, for example: Won, Lost, Ordered, Submitted, etc.

By default, SAP CPQ sends standard values for all listed fields under the Quote Item Mapping section. However, if you are not using a standard value but a custom value for a field, enable the Custom Value toggle and create a formula that will retrieve the custom data.

In SAP CPQ, inference mapping is a part of the integration with SAP CX AI. Inference Mapping is the process of mapping the relevant fields on SAP CX AI back to SAP CPQ after the AI system has analyzed the data and generated insights or recommendations.

Like ingestion mapping, inference mapping is a critical step to ensure that the data returned from SAP CX AI can be interpreted and applied correctly in SAP CPQ. It involves the careful mapping of fields between the two systems to create a seamless flow of data.

The inference phase occurs immediately before receiving recommendations for a quote when CPQ sends a request with data defined from the mapping to CX AI. By default, CPQ sends standard values for all listed fields under the Quote Item Mapping section. However, if you are not using a standard value but a custom value for a field, enable the Custom Value toggle and create a formula that will retrieve the custom data.

Cross-sell CX AI's purpose is to increase order value by suggesting relevant additional products while the customer is making a purchase. By learning from historical transaction data and customer behavior, the AI predicts which products they are likely to buy together.

The cross-sell CX AI generates real-time cross-sell recommendations based on the following general assumptions:

• Customers who purchased X, Y, Z also tend to purchase A, B, C. For example, customers who purchased laptops also tend to purchase laptop cases.
• Customers with X, Y, Z in their quote also purchased A, B, C. For example, customers with printers in their cart also purchased ink and paper.
• Product X is most popular among customers in a specific segment. For example, mid-tier insurance is popular among customers in the SMB financial segment.

The cross-sell takes in quote and opportunity data to train multiple product recommendation models based on the customer's purchase history, the customer's current quote contents, and the most popular product by similar customers in the same segment. CX AI for SAP CPQ uses the models to generate recommendations.

At prediction time, SAP CPQ sends the content of a quote and a customer ID to the prediction end point. CX AI then returns a number of cross-sell recommendations that can be provided to the SAP CPQ end user.

In some cases, due to insufficient data quality or quantity, the recommendation system cannot make a tailored cross-sell recommendation at a high level of confidence. In this case, CX AI provides a recommendation of the most popular product among customers who are most similar to the customer. Even in cases where the cross-sell model quality is low, the recommendation system can still provide a personalized cross-sell recommendation.

Product and pricing recommendations are based on a customer's previous sales data stored in SAP CPQ and on the data of CRM opportunities associated with SAP CPQ quotes.

CX AI recommends only products that are currently not included in the quote.

When recommendations are available for a quote, the Recommendations button displayed as a yellow light bulb is active. Select this button to open a list of recommendations with the following options:

• Add Product: Add a simple product to a quote. The product has a predefined list price. After adding a recommended product to the quote, the product is added at the end of the list of items and is no longer recommended by the system.
• Change Discount: Apply a discount to a product that is already in the quote. CX AI provides recommendations only for the NRC discount. An icon displays next to the recommended discounts for the quote items.
• Decline: Reject a recommendation.

When there are no recommendations to display, the Recommendations will be greyed out. The system learns from the user's activity and adjusts future recommendations accordingly.

The pricing CX AI generates real-time recommendations for pricing markups and discounts. This use case considers quote and opportunity data to train a predictive model that recommends the best discount or markup for a particular product for a specific quantity. The recommendations are customized for the sales representative and the customer and optimize the likelihood of closing a deal.

Pricing recommendations are only supported for simple (non-configured) products and are available only for non-recurrent prices.

As products are added to a quote either manually or by accepting a cross-sell recommendation, SAP CPQ presents recommended pricing from CX AI as a discount of up to -100% or a markup up to +100%.