Similar items based on current item
How does similar items based on current item work?
When you select this recommendation strategy for a touchpoint that’s displayed on an item/product page, it checks what exact item the visitor is looking at. All the item’s attributes, like name, description, category, price, size, and others, are used in calculating the recommendation.
A matching algorithm scans all different “terms” in these attributes and compares them to the items in the content collection linked to the touchpoint. The items with the highest match rates are returned and will be displayed in the touchpoint.
When should I use this strategy?
When you simply want to show visitors similar items on item pages. Note that this recommendation strategy doesn’t consider the visitor’s historical behavior.
What are the requirements for using similar items based on current item?
This recommendation strategy only works on item pages on your website and when our tracking script recognizes the item page. Please read this article to check if that’s the case for your website.
Also, make sure you target the touchpoint using this recommendation strategy to only item/product pages.
Example of similar items based on current item
As you can see in the recommendation on the item detail page above, this recommendation strategy uses the “terms” from the title and category (amongst others) to build up a recommendation that displays similar items.
Others also bought on current item
How does others also bought on current item work?
When a visitor is on a product/item page, this recommendation strategy uses the sales data from that item and checks in historical purchase data what other products people also bought alongside the item viewed. The more the item has been bought, the higher the chance is the item will show up in the touchpoint’s recommended products.
After this data is gathered, the algorithm will also look at the visitor’s general profile (history, current URL, etc.) and combine this with the collected sales data to calculate the best recommendations. This strategy also looks into the values of all viewed, added to cart, and purchased content items and weighs the best items based on machine learning. Therefore this strategy will adapt and change served content items during the visitor's session. Over time, recommendations using this algorithm will improve.
When should I use this strategy?
You could see this strategy as a cross-sell strategy on steroids because it looks at products that have been bought by others who also purchased the item the visitor looked at. But not only is this data taken into account, but also really personalized data on the visitor itself. This results in a recommendation that makes total sense for the visitor since it’s related to the item being looked at and their interests.
What are the requirements for using others also bought on current item?
This recommendation strategy only works on item pages on your website and when our tracking script recognizes the item page. Please read this article to check if that’s the case for your website.
Since the recommendation also uses sales data, you need to ensure that the Datatrics conversion script is correctly set up. Use this article to make sure if that’s the case.
Also, make sure you target the touchpoint using this recommendation strategy to only item/product pages.
Example of others also bought on current item
Alternative items based on current item
How does alternative items based on current item work?
This recommendation strategy looks at the current item a visitor is looking at and checks for the attribute “alternative_items” and will only recommend items from this list.
When a recommendation is made, and there are more items on the list than there are to be displayed, the algorithm will check the visitor’s history and recommend the most relevant items.
When should I use this strategy?
When you want full control over the items that should be recommended on specific item/product pages.
What are the requirements for using alternative items based on current item?
This recommendation strategy only works on item pages on your website and when our tracking script recognizes the item page. Please read this article to check if that’s the case for your website.
Next to this requirement, you also need to add an attribute to the item in Datatrics, which contains the items you want to recommend along with the viewed item. This field needs to be called “alternative_items” and it can contain one or more itemIDs in an array:
["123456", "123457", "123458", "123459", "123460", "123461", "123462", "123463", "123464", "123465", "123466", "123467"]
In the touchpoint’s connected content collection, you need to ensure these itemIDs are also present.
Lastly, make sure you target the touchpoint using this recommendation strategy to only item/product pages.