Billabong is an iconic swimwear retailer.  They are part of a collection of surfwear and board sport accessory brands, including DC Shoes, Roxy and Quicksilver. 

Because of their global presence and multiple brands, it is important that merchandising across assets is easy. This made Barilliance an excellent personalization partner.

We started with their US assets. 

After experiencing the increase in engagements, conversions, and revenue, Billabong decided to roll out the proven solutions across both geographic and brand lines.

This case study focuses primarily on how Billabong makes use of Barilliance's AI and machine learning. 

Billabong’s personalization strategy presents relevant offers during each step of the customer journey.

As customers go through the various pages, conversions are optimized by fitting the recommendation logic in real time. To do this, Billabong uses Barilliance's AI capabilities. 

Barilliance recognizes which page the customer is on and automatically adapts the recommendation strategy accordingly.

The recommendation logic is optimized based on a) what the AI knows of the customer and b) the context that the customer is in (which page they are viewing). 

The home page logic tree is a straightforward example.

On the home page, there are two possibilities.  A visitor can either be a new prospect, who has never engaged the brand before. Or a visitor can be a returning customer, who has previous interactions with either online or offline channels.

By default, Barilliance handles each of these instances differently.

Billabong is able to specify what metric they would like optimized and allow the recommendation engine to do the rest.

I want to walk through a typical customer journey for a Billabong customer, starting with the product page. 

Billabong makes use of two recommendation widgets on product pages. 

The first widget is placed right below the product details. Here they elect to restrict products shown to those that are in the same category. In this case, the product recommendation engine is pulling a combination of top selling, high priced bikini tops. 

When a customer navigates to a different category, the resulting product recommendations change.

Here is an example from Billabong' s shoes category.

These widgets are defined by

In addition to the main product page recommendation widget, Billabong makes use of a second "Recently Viewed" widget. 

Here, product recommendations are dynamically updated based on which products customers have viewed. 

These widgets are defined by:

The next step on the customer's journey is the cart page. 

Here, Billabong is able to offer highly targeted complimentary products. Again, they allow Barilliance's recommendation engine to identify which products are most likely to be bought given the current items in the cart. 

This technique is called collaborative filtering. 

Collaborative filtering can be used to successfully