I've been playing the segmentation game for years, and unless you’ve been living in a cave for the past ten years and haven’t so much as browsed an eCommerce website, you’ve seen product recommendation and segmentation technologies at work a time or ten. Amazon.com, as you probably know, championed the technology of “collaborative filtering” for product affinity models.
Yep, using product recommendation technology for cross-sell, up-sell and website personalization campaigns is nothing new. But I've seen companies not set up the right strategy for executing a truly successful segmentation strategy. And it's OK, it can be daunting. The challenge is making sense of the wealth of analytics and user information, all of which are based on multiple factors, and gleaned from every type of visitor coming to your website. As desperate as you might be to sink your up-selling claws into their shopping carts, missteps will kill your site’s credibility. So get it right the first time.
There are three essential data methodologies for setting up successful, revenue-driving segmentation and recommendation campaigns, that will ensure you “know before you show”:
1. Click Behavior
Simply counting clicks to understand visitor behavior had its brief fame in the early days of web analytics. But the days of rudimentary analytics are long gone. With the resources available today, it’s time to move on and get a little more personal. It’s a deep analysis of click behaviors that gives a more insightful picture of your consumers.
There are 4 key click behaviors that are essential to understanding your customers and how to reach them:
- Viewed this, also viewed that
- Viewed this, bought that
- Most viewed
- Rated this, rated that
Not only does click behavior showcase a user’s navigation through your site, but it reveals their product interests and possible segment categories, allowing you to better focus your segmentation and recommendation campaigns based on your users’ interest, product offerings and promotions. You may even discover new segments and behaviors that you didn’t even know existed. A double win!
2. Basket Analysis
There is no better place to plan for and understand your actual consumers than at the check-out process. Analyzing buyer behavior from all your channels (not just the website!) is the most indicative and predictive data set for setting up successful segmentation and recommendations: discover what people actually buy, uncover buying patterns and help influence inventory decisions.
Marrying offline and online data from CRM, web analytics, POS output, call centers and mobile commerce will ensure you have a 360-degree buyer view to begin targeting and tailoring for your consumers. After the all the data nuptials are complete, analyzing behaviors such as “Bought this, Bought That” and “Best Sellers” will have you homing in on the best cross-sell and up-sell opportunities, for all your point-of-sale outlets.
3. Click & Buy
While some users zoom to your site already knowing which product they are going to purchase, there is plenty of data to be had from those who first spend a lot of time clicking around, researching, comparing products, saving items to a wish-list, and finally…finally, making a purchase. Discovering behaviors such as “viewed this, bought that” will reveal both interests and buying behavior. A combination of this click behavior and basket analysis can help you easily pinpoint segments and recommendations to target at the right time in the buyer’s decision-making process, as well as discover which avenue their most likely to complete the sale at.
Hint: Take #1 and #3 and put ‘em together!
I’ll leave you to start gathering and analyzing this goldmine of data you’ve probably been sitting on for quite some time. Brace yourself: it’s going to be an eye-opening discovery session for your entire business. And your customers will (silently) thank you as well, with a lot more loyalty.
