INDIVIDUALIZE - What could be a/b tested first?
- Sylvain Paillard
All the elements of the “What to start with?” list could also be used in a/b testing. The list here-after provides additional Best Practices that can be quickly tried.
Group | Name | Short description | |
|---|---|---|---|
| 1 | Neural Collaborative Filtering for individual recommendations based on the purchases of the customer and leveraging the patterns in the purchase behaviors of other customers | ||
| 2 | Last Purchase > Often Bought After Last Purchase > Collaborative Filtering Bought After | Promote products often bought after the last purchase of the customer (first practice with statistics, second variant with AI Collaborative Filtering) | |
| 3 | Purchases > Often Bought By the Same Customer Purchases > Collaborative Filtering Bought By the Same Customer | Promote products often bought by the same customer (based on the purchase history of the customer) (first practice with statistics, second variant with AI Collaborative Filtering) | |
| 4 | Promote products or content in the same Price Quadrant as previously bought products. | ||
| 5 | Purchases > Defining Tags | Promote products or content with Discovery Tags matching previously bought products. Each product gets a list of Discovery Tags representing the most common search terms used to find a product. Each product gets a list of Defining Tags representing which is a subset of the product attributes values that have been identified as important in the selection of a product. | |
| 6 | Wish-list > Products Wish-list > Products Attributes Wish-list > Products Discovery Tags Wish-list > Products Defining Tags | Promote products
See Discovery and Defining Tags definition above | |
| 7 | Product Views & Engagement Trends | Most viewed or baskested items within different time frames | |
| 8 | Promote products based on the self-learning of the Widget click/display score, either directly on the amount of clicks or on the conversion, average bought value or margin generated by the clicks Display = visual impression detected on the screen of the user before, during and after scrolling on the page |