Release Summary - Nov 2022

Enterprise Dashboard

MVT - Enabled Multiple Values for Each Dimension to Filter

Introduced the ability to filter MVT reports by multiple dimensions to get insights into how visitors perform against different variations so that it can better serve the right experience. This feature enables the user to select multiple values for each dimension and hence provides the capability to report MVT results for multiple combination filtering.

Filters are based on the following use cases:

• Filter to the right visits that experience the test variants or control.
• Filter by Segments or other attributes to determine how different segments perform. For example, the test may show the winners are different for different segments.

-------------------------------------------------------------------------------------------------------------------------------------------------

Added Attribute for Description to be Used for NLP Models

Optimization managers want to configure the description to be used from an attribute already existing in the catalog so that they don't have to load a separate feed file.
On the NLP Configuration page added an option to enter the name of the attribute in the product catalog to be used as the description for the product. Users can either enter the attribute to be used for the description or select the option to provide a separate file and fill in the option.

-------------------------------------------------------------------------------------------------------------------------------------------------

Introduced Total Data by Month for Engagement Report Rec Type Analysis

Optimization managers can now view the engagement analysis day wise as well as month wise so that it helps them analyze daily and monthly for Rec Type Analysis report.

-------------------------------------------------------------------------------------------------------------------------------------------------

Introduced Size Based Personalization

Personalization is about understanding all details of the shopper/user at an individual level. By understanding the search and purchase behavior of the products, AI engine now personalizes the recommendations to the products that are “best fit” to every user, thus improving shopper’s experience and increasing the likelihood of a purchase.

UPS

Updated Handle Columnar Delta Segment in Attribute APIs

Earlier, the attributes API server didn’t support delta segments updates in UPS, we have enhanced the attributes-API service to support the delta segments updates in UPS. Per the new enhancement, the attributes API must process delta rr-segments feed from RCDP systems.

Added New API for Accepting Offline/Omni Channel Events to CXP

Earlier, customers and other systems such as RCDP were using either 'Offline Data Feed' or RecsforPlacement API to send offline/omnichannel transaction data to CXP. The feed approach was not time-bound, and issues were noticed when using RecforPlacement API to send offline events.

Evolved the UPS API to enable external systems (such as RCDP or customer systems) to send offline events which includes orders and returns.

Find 

Applied Assortment Filter to Sub Assortments

Assortments and sub assortments are parent child relation, hence introduced a new URL parameter subAssortmentFq which sends child filters.

Data Engineering

Rollup - Enabled Calculation for Lift and Confidence Level for Dynamic Experience Variations Against Base (Control) Variation

Digital optimization managers can now see how one dynamic experience variation is doing against other variations in terms of lift and confidence level for the key KPIs through statistical analysis (similar to MVT).

( Full Summary TBD)