Overview

This is an introduction to key Best Practices concepts as applied to Advanced Merchandising in Recommend from Algonomy.

Note: Every retail site is different and what works best for one is not guaranteed to translate into gains for another. As such, treat this guidance as general best practices—a foundation for optimization from which site-specific tuning should occur.

Following the guidelines on the pages that follow helps safeguard the performance of Recommend on a retailer's site.

You should understand the logic behind our optimization methodology, and the questions to ask, so that you can adapt and apply as appropriate.

Best Practices

The intent of these best practices is to navigate you through common user experience decisions when setting up Advanced Merchandising for a retail site. In addition to providing specific guidance on the baseline best practice recommendations configuration. It helps you understand the logic behind our decision making, so that you’re prepared to make informed adjustments specific to a retail site.

Understanding Advanced Merchandising

Advanced Merchandising provides a scalable method of managing precise cross-sell and upsell recommendations. It is a manual merchandising method that can be used to recommend compatible products; such as recommending compatible headphones, cases, and other accessories when a customer adds a new cell phone to their cart, or recommending matching shoes, jewelry and purse when a customer adds a dress to their cart. It is important to remember that Advanced Merchandising is a manual merchandising solution used to supplement the behavioral recommendations to address specific merchandiser business requirements.

It can also be used to recommend upsell opportunities; such as recommending progressively higher-end models when a customer is looking at a new television or computer.

Using Category Data and Product Attributes

To make the best use of Advanced Merchandising, quality data is required. But, what does that really mean?

Because Advanced Merchandising rules make use of product attribute data and/or category data, it is paramount that the category structure is well-defined and/or product attribute data is richly-populated and passed in the product feed. With a well-defined category structure, retailers can use Advanced Merchandising to build category-based rules. With strong product attributes that are consistently applied across the catalog, attribute-based rules can be created.

If the retailer does not have a well-defined category structure or strong attribute data, their use of Advanced Merchandising will be confined to manual merchandising at the single product level.

Good/Poor Advanced Merchandising Rules

It is important to remember that Advanced Merchandising rules, like all manual merchandising rules, come with a certain cost in terms of processing time. A few, well-defined rules can improve performance of your Recommendations, but too many rules or poorly-defined rules can have the opposite effect by slowing down the process or overly limiting the products actually returned in the recommendation.

Examples of Well-Defined Advanced Merchandising rules

  • A clothing retailer uses Advanced Merchandising to drive "Complete the Look" recommendations based on product attributes. For example, building an Advanced Merchandising rule based on products in the category "Dresses" that keys on the attribute of "Color" to also recommend a pair of shoes and a jacket or piece of jewelry.

  • An electronics retailer uses Advanced Merchandising rules to drive recommendations when a customer is looking at cell phones. For example, building an Advanced Merchandising ruled based on attributes of a phone model to recommend a matching case or offer an upsell to the next model.

  • An office supply retailer uses Advanced Merchandising to drive recommendations of consumable products related to an item in the cart. For example, building an Advanced Merchandising rule based on the attributes of a printer to recommend compatible ink cartridges, or the attributes of a calculator to recommend batteries.

Examples of Poorly-Defined Advanced Merchandising Rules

  • A furniture retailer uses Advanced Merchandising to force recommendations across existing sub-categories when the customer is viewing a top level category, such as forcing a Folding Table rec, a Conference Table rec, a Desk rec and a Table Parts rec into the four slots of a recommendation set. In this situation, the recommendation is best left to the behavioral engine.

  • A department store uses Advanced Merchandising to force recommendations that are too narrowly-defined, such as forcing brand and color matches in a niche category like luggage. If the customer is looking at purple luggage from Brand X, a rule that forces recommendations to only be populated by other Brand X purple luggage could very well return no recommendations, missing an opportunity to suggest another brand or a similar color.