Midocean had multiple data-driven questions. Firstly, the e-mails sent to customers were static and in need of personalization. Secondly, there were possible improvements to the stocking process.

Solution

By transforming and cleaning the data, we were able to build predictive models. One of these models is segmenting customers into different groups, viewable in a custom build dashboard. A second model predicted product sales, which enables smarter stocking of products.

Result

The employees of Midoceans’ marketing department are now able to match products with customers and vice versa. Also, we optimized the stocking process by combining sales prediction and trends.