Machine Learning Journey of Burberry: Is There Room For Machine Learning in Luxury Goods Industry?

Luxury retail brand Burberry is one of the machine learning pioneers within the luxury goods industry. Burberry intensively uses machine learning algorithms to improve customer service, to fight counterfeit products and to offer personalized shopping experience to the customers.

British luxury retail giant Burberry is one of the pioneers of machine learning in retail industry and the executives are trying to position the brand as an “end-to-end digital business”[1]. I believe machine learning is particularly important for luxury retail brands due to three reasons: to provide personalized and superior customer service, to avoid counterfeiting of brand’s products, and to offer products that are in line with the current trends. First, personalized and superior service has a high importance in luxury goods industry since customers are willing to pay high prices not only for high quality of the products but also for the spoiling experience that luxury brands offer. With the help of machine learning, brands could predict their customers’ preferences by leveraging all the data they can access, improve their online and in-store experiences by offering tailored recommendations, and use social media accounts of customers to communicate with them. Second, many luxury brands are counterfeited very frequently, and this situation could significantly harm the brands by decreasing the customer demand for original products. Machine learning algorithms help brands identify counterfeited product by only uploading photos of the suspected products into applications. Third, creating trendy and on-demand products are at the heart of luxury goods industry and it is very hard to turn the luxury brands around if they fail offering the right products at the right time. Machine learning allows brands to discover which products are trending and design their products accordingly.

Executives of Burberry are working on several important initiatives to transform Burberry into an end-to-end digital enterprise. Burberry is now collecting customer data via loyalty programs, and using both in-store and online data to tailor its offerings to the customers. To fight with counterfeits, Burberry is using image recognition technology provided by Entrupy which enables the company to identify if a product is genuine only by looking at a small piece of suspected product’s picture. Moreover, Burberry salespeople display customer-specific recommendations on their tablets to provide real time recommendations. Burberry also offers a series of cutting-edge experiences to its customers. For example, Burberry is one of the first brands that leveraged Facebook chatbots to let customers know about their new products, to provide customer service, to let them search nearby stores and to purchase new products – as well as to book an Uber ride to the nearest store.[2]Last but not least, Burberry also offers customers to create their own personalized copy of their ads, allowing customers see themselves next to the celebrities. In the medium run, Burberry is planning to leverage machine learning to better their support processes. As David Harris, SVP of IT, mentioned:

“Areas we are looking at include: chat bots in service functions, proactive IT operations, insight from pattern recognition, automated and scenario modelling for planning & logistics, and security & fraud prevention.”[3]

Image: Facebook chatbot of Burberry [4]

In the short run, I recommend Burberry to actively track the behavior of its customers in other platforms: online retailers such as Amazon and Rent the Runway to understand their current shopping behavior, as well as Spotify to get a sense of what they are listening and Yelp to grasp where they have their dinners. I believe that luxury shopping is a comprehensive experience, and customers better respond to the brands that are in line with their lifestyle. By using other data sources, I believe Burberry could better know their customers, develop in-depth understanding of what they really do, and offer the most suitable products to their customers. For example, for the customers who eat at more cozy places and listen to country music, Burberry could offer sneakers and sweatshirts. On the other hand, for the customers who eat at fancy places and love listening to popular party songs, it could be more appropriate to recommend high heels and fancy trench coats. In the medium run, I recommend Burberry to use the customer data in product development. I believe that having a trendsetter position is particularly important for luxury brands, given that one of the main reasons why people choose to buy from a luxury brand is the desire to be associated with a trendy and cool brand.

Two questions: 1) Is using machine learning to streamline support functions in a luxury retailer priority, given that cost is not the most important element of the business? 2) Does customer preference drive luxury products, or do luxury products drive customer preference? If the latter is true, is it possible to predict customer preferences from the historical data? (751 words)

[1]Bernard Marr, “Burberry: How Big Data and AI is driving success in the fashion world,” https://www.bernardmarr.com/default.asp?contentID=1282, accessed November 2018.

[2]Forbes, “The Amazing Ways Burberry Is Using Artificial Intelligence And Big Data To Drive Success,”  https://www.forbes.com/sites/bernardmarr/2017/09/25/the-amazing-ways-burberry-is-using-artificial-intelligence-and-big-data-to-drive-success/#540f075b4f63, , accessed November 2018.

[3]AI Business, “Where are Burberry with AI? Exclusive Interview with David Harris, SVP of IT,” https://aibusiness.com/where-are-burberry-with-ai-exclusive-interview-with-david-harris-svp-of-it/, , accessed November 2018.

[4]Topbots, “Burberry,” https://www.topbots.com/project/burberry-facebook-messenger-chatbot-guide/, , accessed November 2018.

Previous:

Minimizing Decision Fatigue Through Machine Learning @ TripAdvisor

Next:

Ant Financial – Pioneering China Fintech with Machine Learning

4 thoughts on “Machine Learning Journey of Burberry: Is There Room For Machine Learning in Luxury Goods Industry?

  1. Luxury brands are in the latter wave of adopting digital compared to mass consumer goods company. One main reason is that they still value the in-person experience with the customers, given a lot of time it is the experience and superior service make what luxury goods luxury. Luxury brands have long been hesitant taking offline to online. What is your take on the machine learning and digital disruption in terms of in-person consumer experience?

  2. Great article that highlights the advantages of information sharing and targeted marketing in the luxury goods sector. Providing targeted advertisements to specific potential customers, as you highlighted, may risk some damage to brand image and could potentially target the wrong people. Burberry clothes are definitely not for everyone and data should be used to indicate ability and potential to buy as opposed to product preference. Should burberry target all potential consumers with specific products, or should they use their data collection to determine who potential customers and then target with a wide array of products? Offering burberry sweatshirts to people who prefer comfort and country music could prove unprofitable and certain traits may indicate that a person isn’t inclined to be a customer and should be ignored. It is an important balance between customer selection and targeting that needs to be weighed when making this decision.

  3. Loved this! Especially the part on how machine learning may help companies identify counterfeit products! That’s one aspect of machine learning in the fashion industry I had never thought about previously. The second question is particularly interesting. Although I’m not a fashion guru, I believe luxury products drive customer preference. This makes me wonder if Burberry dilutes its brand value by pushing its individual products on customers. Machine learning does provide some important insights on customer preferences but perhaps more subtle advertising may be more line with Burberry’s image. For example, instead of pushing particular products like sweatshirts to customers, Burberry could just customers a simple alert (for example, “Check out our new fall collection”). This fall collection would include products that were created using machine learning insights. This approach would also help Burberry avoid being perceived as pushing products on customers.

  4. Great article Bilge! I really enjoyed reading about how Burberry could use machine-learning to grow its business. This topic is very relevant to my topic of how H&M is using machine-learning to turn its company around. Since H&M is a fast-fashion retailer, its metric for success is successfully predicting a trend. On the other hand, I would argue that Burberry, a renowned fashion-house, is responsible for CREATING the trend (not predicting one). Consequently, I would recommend that Burberry not use machine learning to inform its designs because it is not only a risk to their brand but also they can not keep up with the quick pace of manufacturing that trend prediction and execution requires. Lastly, I’m not a big fan of Burberry using chat bots. As you noted, I think Burberry is a luxury brand and needs to maintain an elevated experience. In the current state, I find chat bots to be quite “mechanical” – which is the opposite of a warm, tailored experience a luxury customer would require.

Leave a comment