AI Recommendations by Loop54
What does this function offer?
Recommendations are an important tool for offering personalized product recommendations online. They help find what the customer is looking for and provide specific suggested products based on their preferences and wishes.
Recommended Products
The use of purchase history and personal preferences enables the presentation of specially selected products that are relevant to each individual visitor.
The use of purchase history and personal preferences enables the presentation of specially selected products that are relevant to each individual visitor.
Similar Products
Our AI algorithm maps relationships between products and recognizes which other products match the user's intentions. This enables the display of a wider range of relevant products and inspiring alternatives.
Manual product recommendations
Manual recommendations will be available as before and will be given priority during sorting. If manual recommendations are missing, they will be added automatically and personalized.
What do I have to do?
Once you have chosen the Loop54 Engine, we take care of the technical part of the integration. The engine effectively needs no manual input or configurations and starts working and learning immediately.
You can find the control panel under Modules → Loop54 Settings in the FACT-Finder backend.
There you can set the maximum number of displayed and used recommendations, e.g. 10 recommendations.
Manual Recommendations
Manual recommendations (flagged) will be available as before and will be preferred in the order. If manual recommendations are missing, they will be automatically supplemented in a personalized manner, as can be seen in this example of "Mandalorian, Star Wars".
You can find more information about Manual Recommendations here.
What do I need to look out for?
During setup, we activate this feature by default, since a Loop54 engine must be set up at least for one channel.
| Please note the following: |
|---|
| In contrast to Personalization, the Recommendation format includes master items and variants. Currently, as with manual recommendations, a possible variant is randomly selected and issued as a recommendation. For example, an "M" shirt may be recommended for an "XXL" sweater. |
| Similar products are not based on sales information, but on "fields", e.g. similar items by color or product group within the product catalog. Again, variants are randomly selected if there are any at that level. |
| Blacklist recommendations also only work based on the MasterID. |
| The Recommendations page and its Matrix subpage are no longer available. |
| Manual Recommendations will be available as before and will be prioritized in search results. If manual recommendations are missing, they will be automatically replaced by personalized results. |