Ecommerce Personalization for the Mid-Market Retailer Part II

Functional Requirements of Ecommerce Personalization

Part II of our post on ecommerce personalization. Here’s a link to Part I.

Deciding what personalization tool is right for you as a retailer is dependent on requirements such as your industry, how many products are in your catalog, customer life cycle, level of customization in marketing messaging, customer sophistication, and time period from lead to sale.  How you answer these questions will help you identify the right balance from a personalization tool that’s right for your needs.  Personalization tools offer features that can be categorized into three categories: versioning, cross-selling, and interactive filtering.

1. Versioning tools personalize an experience by first defining segments of consumers and then serving up different iterations of key pages of Web sites, for example showing different versions of a home page to consumers. Versioning tools are useful to marketers who want to carefully control their messaging.

2. Cross-sells. Simple cross-sells take implicit data and simply place relevant product “adjacencies” on a product details page. Cross-sells are low complexity, inexpensive, easy-to-integrate and simple solutions that help to automate tedious merchandising processes. Small to midsize retailers are the most active customers of cross-selling tools. Advanced cross-sells have the capability to push suggestions to other parts of a site or company such as onsite, email programs, POS systems or call centers.

3. Interactive filtering solutions ask consumers for specific inputs, posing a series of questions and then matching responses based on their preferences. This allows consumers to tell the retailer what sort of information they want.

If retailers have extensive resources, both in headcount and budget, to dedicate to personalization engines they should consider solutions with versioning tools and advanced cross-selling tools.  If retailers have an extremely broad or complex assortment of products they should consider using solutions with simple cross-selling, advanced cross-selling and interactive filtering tools.  If retailers need to exert close control / input over content displayed on its site it should consider solutions with versioning tools, advanced cross-selling and interactive filtering tools.

Typical Challenges Retailers Face

The “Locked Loop” refers to the challenge of how to incorporate the long tail in product recommendations. There is a higher probability that displayed products will get purchased, purchased products will get recommended over non-purchased products, and recommended products get recommended again causing less popular products not to be displayed. The goal with automation is to provide relevant content to site users but don’t lose the surprise and delight factor customers seek during their shopping experience. Be sure to audit your product recommendations regardless of the business rules you use.

The “Cold Start” is the period of time from initially setting up the recommendation engine to collecting enough data to make useful recommendations. Cold starts can be mitigated by utilizing historical customer behavior data as an initial input getting a quicker return on their investment in the tool than those who wait to use future site users’ aggregated behavior data.

What’s the ROI to Personalize?

Retailers who use personalization engines historically have seen measurable lifts in their performance metrics including an increase in units per transaction, increased AOV, more time spent on the site exposing customers to more content and overall revenue lift.  According to a study 15% of consumers purchased a product recommended to them, and of those shoppers 62% indicated that recommendations were very useful in making purchase decisions. Automating merchandising also frees up significant time for employees to engage in other efforts, historically spending 75% less time than if they were to merchandise a site manually.  A Forrester study of 30 top internet retailers (sales greater than $100MM) who used third party personalization tools saw marked increases in site performance, including

  • 65% increase in new customers
  • 5% higher AOV from new customers who clicked on recommendations and purchased
  • 35% increase in repeat customers
  • 10% higher AOV of repeat orders who clicked on recommendations and purchased
  • 5% direct incremental lift

Eye tracking studies and clickstream studies show above the fold items have a higher probability of getting clicked so keep recommendations close to the hottest paths of the click stream.  Also, consider exposing the logic why you are showing shoppers recommendations through verbiage.  Don’t just tell her, “You Might Also Like” when you could be saying, “Shoppers Ultimately Bought These Items When Viewing Product X”.


There are now tools available to mid-market internet retailers that can be integrated with Magento Enterprise.  For retailers who are just getting started in product recommendations, consider utilizing Personal Merchant by Predictive Intent, or the Amadesa Personalization Suite.  For other retailers who have more experience around utilizing historical customer data and product recommendations, consider looking into the RichRecs Express tool by Rich Relevance.  Large internet retailers can take advantage of more robust tools like Rich Relevance’s RichRecs Enterprise or Adobe’s Omniture Test & Target.  Because personalization is not yet widely accepted retailers who employ it can distinctly differentiate themselves to consumers where other competitors lack a personalized shopping experience.