Our prediction engine uses deep segment analysis. This looks for deviations from normal, or bias between customers that purchased product X and those that did not. Unlike cross-sell analysis which is typically used for recommendations, our approach generates relationships between products by essentially filtering out the common purchases (purchases that are typical across your store but not necessarily related to product X). 100% of the customer base and 100% of transactions are analyzed when generating relationships.

Using your store’s live and historical data, recommended product suggestions are included in mailouts. When you drag and drop a "recommended products" container into your template, our recommendations engine will automatically generate a product when the email is sent to the customer. 

Recommended products are the products your customer is most likely to purchase next based on past purchases.  

You can customize the container, adding multiple recommended products which can also automatically include the product title, description and price. You just drop the container and pick your information, and then we do the rest of the work :)


For stores that are new and do not have a lot of orders (or any orders), you will need to insert static product images into your emails and will not be able to use our Product Suggestions initially.


Recommendations work by looking at what someone has bought and then compares them to similar customers (that bought the same thing) and unlike them (bought different things) to put in product recommendations that are individualized.

If there is nothing to be recommended (for example in this case where no one has ever bought anything on the store yet or very few have) we will have nothing to recommend to them. If there are product recommendation images in the email, but nothing to recommend we won't send the email and the test you receive will have blanks where the product recommendations were placed.

Did this answer your question?