Recommending products for each customer or finding customers who are likely to be interested in a particular product
All of your customers will have different wants and tastes, so you will sell more if you promote the right product for each customer. Using deep learning – which is type of AI – we analyze the following factors:
- Product purchase history – if people who buy one product regularly buy a second product then that is an indication of a pattern.
- Product collections – buying a particular category of product means that the customer is interested in that type of product.
- Price – customers often prefer products at a particular price range.
- Predicted gender – men and women frequently buy different products.
- Country – in fashion, food and drink tastes are different between countries.
- Area – regions within countries also have differing tastes.
- Mosaic Group and Type (the Mosaic subscription add-on is required)
Not all of these factors will be relevant for every brand but if a pattern does emerge it will be taken into account. Recommended products are produced for each customer, with you can then view on the Contact record.
We can also do this in reverse, so if you have a product that you want to sell (perhaps because you have an overstock or you have products reaching the end of their life) then you can create a segment of your customers who are most likely to want to buy this product.
Minimum number of orders
The more orders our AI has to look at the better it will work, as patterns will become clearer with larger numbers of orders. As a minimum you must have:
- At least 1,000 orders in your shop; and
- At least 25 customers who have ordered 2 (or more) products.
If you have less than these limits you will not be able to see the product recommendation feature.
Sending emails with recommended products
If you are sending an email campaign and want each customer to see the product they are most likely to buy, just drag a Smart Products block into your email and then choose Recommended from the product selection box.
Recommendations for a particular customer
You might want to see the product recommendations for a particular customer. Maybe the customer has emailed you or perhaps the customer has bought something that is out of stock and you need to suggest an alternative.
Just go into Contacts, then view a customer who has bought from you to see the recommendations. Note that recommendations will not be available for the newest customers.
Finding customers interested in a particular product
If you have a particular product that you want to sell, you can create a segment of customers who are likely to be interested in buying it.
Just go into Contacts, then Segments, then click on Create a new segment. Then select Recommended product is from the menu.
This can be combined with other conditions in the normal way, so for example, you could find customers who are a particular size for clothing or customers in a particular country.
Preventing products from being recommended
There might be certain products or categories that you never want to appear in recommendations. You can exclude products by going to the Brand Settings from the menu in the top right corner, then scrolling down to the Product management section.