Build an ecommerce personalization strategy
1. Use consumer research knowledge you already have from your CRO, and internal conversations to identify possible attributes for segmentation and potential segments.
Get help from product experts, buying departments, and analysts to look for broad, data-backed hypotheses about the customer base. You can do cluster analysis on purchased product categories, brands or percentage discount using orders data and check the results against your proposed segments. Hypothesized segments that do not pan out in the data should be ignored.
2. Make a list of all the segments and associated experiences you want to assign to them.
For example, discount sensitive customers, customers preferring high-end brands, and hobbyists, with their respective personalized experiences. For example, changed copy, workflow, layout or features, popups and banners, background pictures, countdowns, and product recommendations.
3. Give a score to each segment on your list and prioritize which to target based on expected value and ease of implementation.
You can base it on the size of the segment, the resources and time needed to build the relevant experience, and the magnitude of the uplift you anticipate from it. Give a higher score to the higher segment sizes, to the higher expected values and to the lower requirements, then add all these up. See which segments have the highest scores and those are the ones that have the biggest impact and require the lowest effort.
4. Email a personalized newsletter to half of the members of a segment to test that segment.
Email the default newsletter to the other part of the segment and general audience but the variation to only part of it. You can test the segment in other channels such as advertisement displaying and product sort order, too. The most simple channels are the ones not suffering from caching issues, users logging out, and other events that make test analysis difficult.
5. Compare revenues for those audiences to check if the experience in the chosen channel yielded an increase for the targeted segment.
With additional resources, all segments can be tested through emails, where three segments and three emails are fully combined to create a 3×3 test in the email.
6. Check if other channels work as well using the segments, if the initial test showed an impact.
For example, you can test sorting the product display on the website based on the segments.
7. Combine segments to make the experience for the customer more personal, if multiple segments have shown to work.
For example, combinations of product category segments and brand segments could work together. Use a scalable content hub or digital asset management tool like Bynder or Nuxeo to scale content creation together with personalization.
8. Target users based on recent meaningful-for-its-segment behavioral data, like a product category looked by the user.
For example, if a premium brands user looks at four BBQ grills, of which two are premium brands and two cheaper brands, the primary target should be premium brand BBQs and accessories. This targeting can be applied through multiple channels too.
9. Set up a recommendation algorithm.
For example, a set of recommended products based on what other people viewed within the same segment.