SHISEIDO: Leveraging the Power of Machine Learning to Boost E-commerce Performance

Challenge

Personalizing the Customer Experience with First-Party Data

Shiseido is an iconic beauty company known for its skin care, hair care, cosmetic and fragrance expertise. Based in Tokyo, Japan, the group operates in approximately 120 countries and regions around the world.

Shiseido has proven itself to be a leading brand in its commitment to global digital transformation at the enterprise level. A core tenant of this transformation is e-commerce performance. It’s no secret that first-party data yields immense value for brands looking to gain increased insight into online customer behaviors, but how can brands effectively collect and harness first-party data for enhanced customer personalization? 

Approach

Customizing Machine Learning to Work for Your Brand

By leveraging fifty-five’s data expertise and machine learning capabilities, Shiseido implemented a framework for defining potential consumer sales value and predicting customer lifetime value, ultimately impacting the efficient use of marketing spend by automating the process for ongoing optimization.

fifty-five analyzed Google Analytics (GA) data to discover the behavioral pattern of users across Shiseido's website. Then, using Cloud Machine Learning (pLTV) with BigQuery, fifty-five predicted the lifetime value of users based on e-commerce interactions.

fifty-five ingested these results into GA for media and website activation. The team successfully automated the model to rerun and intermittently export the results to GA for further analysis.

Results

Nurturing Customers for Life with Meaningful Messaging

The results speak for themselves. fifty-five’s machine learning implementation successfully segmented low and high lifetime value customers and identified conversion rate optimization use cases.

Shiseido continues to collaborate with fifty-five as a long-term data partner with additional planning for future e-commerce performance strategies.

x1.27 conversion rate

24.8% decrease in cost per conversion

x2.2 return on advertising spend

x2.7 revenue per user