Merchandising variables: Understanding a Retail Fundamental

Merchandising variables are fundamental to any successful retail web analytics implementation. Adam Greco has a great post about the technical details of how to set Merchandising variables (with great follow-up comments from our own Kevin Willeitner).

For this post, I wanted to back-up and give an overview of why merchandising variables are so important to retail merchants, and what insights you can get for site optimization.

Merchandising variables can be drive important insights into your customers’ behavior. They can help answer key questions that you cannot get from any other internal reporting tools. For example:

• How much revenue is being driven by internal search vs. traditional browsing (or other methods)?
• What web categories are driving the most revenue? Top categories for a given product (if that product is shown in multiple categories)?
• What site features are customers using (e.g. zoom, alt views, size filters, refinements)?

The challenge with traditional variable attribution is that a customer may have multiple products in an order, with multiple influences for each product. So what features get credit for a given order? Merchandising variables give us the answer: Give credit at the product level to the features and functionality that drove a customer to each product.

Let’s look at an example: A customer views and purchases 4 products on your site. For each product they purchased them from 4 different web categories, and used multiple different site features for each product. How do you allocate the $135 order total? Merchandising variables allocate the revenue at the product level, so each category gets the credit for their product.

https://blog.adobe.com/media_071fc3ea53d1a7c345b0df19d23f58d8e5ad06c4.gif

For this example, I created a sample “matrix” of site features they used for each product (which are custom for every implementation). Each of these columns in green represents a possible merchandising variable that is set when the customer views each product. An analyst can now run these reports individually and understand how that specific feature contributed to site revenue. How much revenue is driven by zoom or alt views? What is the most popular finding method?

https://blog.adobe.com/media_58ed867dac751d42e3d66d0b389b8695bb9e970f.gif

For example, an analyst can run the “Web Category” report and see the revenue breakdown for products purchased from those categories. The site merchants that created these categories use this information to understand category performance, and optimize their them for maximum revenue.

https://blog.adobe.com/media_a9b8b3572ccf31507dceb43d5002c5d3f51e6ea9.gif

Additionally, an analyst can use this report to understand the product conversion rate (look-to-buy ratio), to understand customer interaction with products in that category. In this case, the “Spring Looks” category has a lower conversion rate than the others, so the merchant may want to review the product selection in that category. Also, an analyst can look at a specific product (or product classification) and break down by the web category to see where customers are buying a specific product.

Another report in this example would be the “zoom” report: How much revenue is directly driven from customers who used zoom before purchasing a product? What types of products use zoom more than others? In the example below, an analyst could break down the Sweaters Product Family (a classification of products) by Zoom Usage to see how that feature is used for Sweaters.

https://blog.adobe.com/media_04b2fcb7b9893077e5ada63b964e5dd3b41840f2.gif

Common uses for merchandising variables include any situation where you want to track revenue for product-specific features, such as web categories, product finding methods, search keywords, zoom and alt view tracking, size filtering and refinements.

To summarize, here is what Merchandising variables provide for the retail web analyst:

• Ability to isolate “Direct” product revenue driven by the features and functionality on the site
• Ability to “break down” a merchandising evar to see the products that contributed to the revenue for that feature
• Inversely, “break down” a product to see what features drove the demand for that product

Implementation

Merchandising variable implementations can be a little tricky, so be sure to consult with Adobe Consulting if you plan to take advantage of this fundamental feature of SiteCatalyst.

In a future post, I will list some tips for implementing merchandising variables, and caveats for reporting and analysis.

Comments

I would love to hear your feedback or questions. Thanks for reading.