Publishers of technology books, eBooks, and videos for creative people

Home > Articles > Web Design & Development > PHP/MySQL/Scripting

Building an E-Commerce Site with PHP: Making Product Recommendations

  • Print
  • + Share This
Just as you’re more likely to watch a movie that comes recommended, recommending products to customers is a great way to increase an e-commerce site’s sales. In this article, Larry Ullman walks you through the logic required, and the decisions that need to be made, to implement such a system.
From the author of

The whole point of an e-commerce site is to get customers to make purchases. Partly this means a site needs to remove barriers that might make a customer rethink his or her decision (e.g., the site must have a good user interface, must implement sound security, and so forth). But, more importantly, there have to be items the customer wants to buy in the first place. While a functional browse or search system can help customers find what they’re looking for, being able to provide knowledgeable recommendations can go a long way towards increasing sales.

Recommendation Types

There are two broad types of product recommendations:

  • Upselling: Recommending similar products that are better and more expensive.
  • Cross-selling: Recommending related products that the customer might want in addition to the original, intended product.

Which type of recommendations your e-commerce site is capable of making depends upon what you’re selling. Cross-selling is the most universal: Almost anything can have a recommended cross-sell. If you’re selling coffee, you might recommend biscotti or mugs. If you’re selling music, you might recommend similar artists. Upselling is not as universal: If the site sells custom quilts, the only possible upsell is a more expensive quilt. But the logic for upselling is more straightforward: What is something better? If the customer has a 2 lb. bag of coffee in her cart, recommend she buy a 5 lb. bag instead.

Regardless of what type of recommendations the site will make, you’ll need to determine how appropriate recommendations will be defined[md]i.e., how will the site actually know to recommend product B but not product C?

  • + Share This
  • 🔖 Save To Your Account