Merkle and Leen Bakker have built a central data hub that puts the customer's style preference first. The omnichannel customer experience can now be enriched with a personalized selection of products and content resulting in higher relevance and activation of customers and potential customers.

Leen Bakker’s goal is for everyone to furnish their home cozily, affordably and in their own style. There are two basic principles when it comes to interior shopping: First, you can’t argue about taste. Products appeal to customers if they match that customer's interior. Secondly, there is only one chance for a good first impression, especially online. The starting point for this case was to develop a method to identify style preferences of customers and prospects and classify the assortment into the appropriate style group. 

There were two major challenges: first, to be able to determine the style preferences of customers and prospects of Leen Bakker and second, to classify all 15,000 products of Leen Bakker with the correct style group. Both are needed to show the right products, in the right style to customers and prospects. Taste-driven marketing, so to speak. This requires being able to predict taste at the customer level and classify products by taste.

Two projects were launched to link customer preferences and products into a relevant, personalized message.

- Style recognition for existing and new customers
- Product classification to show the right products

Leen's brain

In collaboration with Leen Bakker, Merkle built a central data hub that puts the customer's style preference first. Through this data hub, Leen Bakker can recognize which style best suits an existing or new customer.

Within Leen's Brain, a Google Cloud Platform CDP, there are two different data sets that take center stage: at the customer and product level. At the customer level, we combine online user data (website traffic, purchase behavior) and offline user data (offline purchase behavior) into a Single Customer View (SCV) in the data hub, a modular CDP built in the Google Cloud Platform. By combining purchase data with labeled item data, it is possible to determine style preference for already known customers and predict style preference for new customers. Because Leen Bakker stores all customer data centrally in its own modular CDP, Leen Bakker can use this data at every stage of the funnel to enrich the omnichannel experience with personalized selection of products and content. This solution is fully future proof in the post-cookie era.

Enrichment and validation of style preferences through email

To gain even more insights on style preference, an email campaign was set up to solicit style preferences directly. In this way, style preferences of existing customers can be validated, but also the style preference of prospects can be identified. Expressed preference of style is a strong predictor of style preference.

Through an omnichannel campaign, customers and prospects were sent to the campaign asking about their favorite home style (style preference), with the goal of lead generation and profile enrichment. In collaboration with the concept stylist, marketing and category management, a mood board was developed for each style preference for each room. Based on the data collected and marketing automation, several personalized emails were triggered. Each email was targeted to one specific space where subject lines and email content were personalized to style preference of the customer/prospect. Leen's Brain then helped by employing machine learning to classify the products by style group.

The segments and look- a like audiences are used in online marketing channels, both paid (SEA, Display, Video and Social) and owned (direct mail and email), combining style preference and the perfect product fit. 

By storing customer information and product information in one central location, Leen Bakker reaches customers with a personalized selection of products and content. And activates these customers and reach potential customers who show the same shopping behavior, thanks to look-alike target groups. This way we connect to the customer's style preference early in the omnichannel experience and increase the relevance of Leen Bakker for each customer, resulting in an increase in performance.&

By using the Single Customer View, it is possible for Leen Bakker to advertise for competitive generic keywords in a relevant way to better reach potential customers. The results of personalization by style have been spectacular. Since then, there has been strong growth in clicks, conversion and ROAS within paid search campaigns and a decrease in costs.