Another new model has been added to Configurable Strategies, Bought Together in Order. With the new bought together model it looks only at products that were included in the same order, while Bought Together by User considers all products purchased by a user in the last 75 days. In addition, you can now include the seed product with the results of a strategy. This is often used to represent product bundles with the seed and products that are often bought together.
To improve collaboration with merchandising rules the following fields have been added to Boost Rules, Restriction Rules, Manual Merchandising Rules, Strategy Rules, and Configurable Strategies:
- Created By
- Create Date
- Last Updated By
- Last Update Date
To view these in the list you can click on the list icon when hovering over a rule.
The Visual AI Complete the Look model has been updated to accept merchandising feedback to improve the training for individual sites. You can choose product recommendations to be used with a seed influence the model. The product recommendations are then curated for the seed product, but also the model is influenced to affect product recommendations for other seeds as well.
Experience Optimizer is an algorithm that decides which content or strategy to play for a user based upon a predictive algorithm that is configured to optimize for a specific metric, such as click through rate. The predictive algorithm uses visitor input to identify patterns of behavior that will influence which content or strategy to play. To be able to apply this algorithm in different types of online experiences, we've opened up the ability to configure the input used about the user for making its decisions. For example, you could optimize only based upon the placement, or combine the placement with another attribute such as their segment.