1. Your Product Recommendations asset will be populated with products that reflect the behaviour of your website visitors. That behaviour is your asset's logic:

Machine learning-based assets will include products according to a combination of potential and actual purchase behaviour of all the visitors of your website.

Crowdsourced assets will look at specific behaviours (browsed, carted or purchased) of all visitors in your website and show products with most interaction during the last three (3) hours.

Personalised assets display those products that the Profile viewing the asset has browsed, carted or purchased, ordered by most recent.



A fallback source will act as a backup whenever there isn't enough visitor data to select the products to be displayed in your Product Recommendations asset. They can complete an asset that is missing one or more products, or even a full asset if there isn't any data to support the primary logic.

The primary logic is the one you've selected above.

The primary fallback source will be the first to be considered when in need of more products for your asset, due to your primary logic not having enough data.

The secondary fallback source is considered whenever the primary fallback still can't provide the data needed to fill the Product Recommendations asset.

It's important that you consider how these three interact with each other. Choosing a fallback source that is more limited or specific than the primary logic may cause your Product Recommendations asset to lack enough products to display. The fallback sources should be more broad than the primary logic.


2. Choose a primary fallback source.


3. Choose a secondary fallback source.


4. Click Next to proceed.


Next step

  • Creating a Product Recommendation: Design

Did this answer your question?