A featured contribution from Leadership Perspectives, a curated forum for enterprise technology leaders, nominated by our subscribers and vetted by the CIOApplications Editorial Board.
Lamps Plus
Website Personalization: Potential Segmentations and Types of Personalization
Due diligence on website analytics and CRM data is required before embarking on a website personalization strategy. Personalization also requires hypotheses, testing, and analysis before scaling any tactic to all website visitors. More advanced personalization requires significant effort to design, code, and QA, so it is also critical to focus on strategies that reach a large audience and make the most impact on the business.
Potential Segmentations
Personalization starts with understanding website and CRM data to group visitors based on the segments and/or clusters that will make the most impact on a business. Since each business is unique, what works for one company may not work for another. For companies with limited resources, it is important to prioritize segmentation tactics based on audience size and the impact on the business.
Examples of segments to consider include:
1. Known vs. anonymous visitors.
2. Known repeat visitors and customers.
3. Customer clustering using predictive analytics (e.g., lifetime value, active, one-time buyer, likely to churn).
4. Product attribute affinity based on preference or predictive analytics (e.g., category, style, finish, price point, sales items).
5. Website browse behavior and purchase history.
6. Geographic location, IP address, or weather.
7. Type of device used.
8. Traffic source (e.g., social, search, affiliate) and referring URLs.
9. URL parameters to identify campaign and creative used.
10. CRM (e.g., consumer vs. trade, new homeowner, assigned sales manager).
For companies with limited resources, it is important to prioritize segmentation tactics based on audience size and the impact on the business
Types of Personalization
Deciding which website personalization tactic to use depends on technology availability and resources required to implement. Here are types of website personalization tactics ranked from easiest to implement to hardest based on these two parameters:
• Recommendation Widgets
Recommendation widgets are the most prevalent form of personalization. Widgets typically recommend products, services, or content based on user browse and purchase history. These widgets typically feature related items, items often bought together, abandoned cart items, and previously viewed items. The business rules for each widget should be different based on the customer purchase journey on site. Some use business rules to prioritize higher margin items, best sellers, or a new assortment. The best practice is to test business rules, widget locations, and message call outs regularly. Many technology vendors for product recommendations have been in business for at least ten years and are sophisticated at a reasonable cost. Implementing recommendation widgets is relatively easy.
• Graphic, Promotional Offer and Layout Personalization
Graphic, promotional offer, and layout personalization requires analysis, hypothesis, and AB testing. It also requires more graphic design, web dev, and QA resources, particularly for a website with a large catalog that still wants to maintain consistent branding with so many dynamic graphics on a page. This personalization tactic allows websites to feature relevant hero images, marketing messages, and prioritize feature categories based on style/finish affinity on the home page and landing pages. For example, if a visitor has shown a stronger interest in browsing how-to articles, watching videos or reviewing inspiration images (e.g., room scenes), they are more likely to see the same when they come back to various sections of the site. These types of personalization tactics are more advanced. Rolling out in phases and performing sufficient levels of AB testing may likely be a better approach. A business should also consider ongoing efforts and how they relate to the business impact.
• On-Site Search, Product Sorting and Navigation Menu Personalization
For e-commerce sites with large catalogs, significant thought goes into what is displayed when a visitor performs an on-site search or goes to a category sort. Most visitors browse a relatively small number of pages before they make a purchase. Thus, the display sequence of a product sort that provides relevant results can move the needle significantly on conversion rate and value. This capability tends to be built in-house and utilizes various algorithms to decide on the default display sequence. Companies have also attempted to display a dynamic navigation menu based on browser behavior. Taking browse history, engagement, and purchase history into account to modify the sort from the on-site search and sort pages for personalization will be the next challenge for e-commerce sites.
Technology has made personalization an effective way to deliver unique, impactful content to users. Knowing how much to implement for personalization is the most impactful way to lead to success.
