Fortune Coffee uses machine learning for optimal customer experience

Many people cannot live a day without it: coffee! Besides that, the black gold is a connector. We drink cup after cup to get to know each other and build relationships.

Together, 30 Fortune Coffee entrepreneurs make sure their customers always serve a delicious fresh cup of coffee to their employees and guests. This coffee supplier not only wants to build customer relations with coffee, but also really add something to the experience. Together with CRM Partners, a major digitization effort was made in 2015 to take the personal experience to a higher level. At the beginning of 2017, it was time for the company to further expand its leading position in customer experience, this time with machine learning. Matthijs Wiegant, marketeer at Fortune Coffee, tells more about this.

Everything for an optimal customer experience

“In recent years, we have worked together with CRM Partners on an optimal service experience for our customers. As a result, we have built up a good relationship. We have taken new steps together and got to know more about machine learning. The digitization process of 2015 has been an important factor in this. At Fortune Coffee, we are always searching for better ways to meet the needs of the customer. We like to know the customer personally and surprise him or her. CRM Partners shares this vision and is truly a partner in the technology that is needed to give our customers an optimal customer experience”, Wiegant says.

How it all happened

After the digitization process, Fortune Coffee was ready for the next step. Wiegant: “In order to keep providing optimal service and remain a leader in Customer Engagement in our market, a new challenge was welcome. At a CRM Partners event, we talked about the application of new technologies such as machine learning, and we were struck by its endless possibilities. After a brainstorm about the various application possibilities, it became clear that this technology can make a big difference in customer experience. Thanks to the innovative nature of Fortune Coffee, nothing stood in the way of starting a machine-learning project. We focused on the most practical idea that benefits both our customers and entrepreneurs: the offer. Before we used machine learning, it was mainly about the knowledge of our entrepreneurs. The problem with that is that when the organization of an entrepreneur grows and more staff gets involved, these new employees do not (yet) have the knowledge to make an appropriate recommendation. Moreover, with machine learning, the entrepreneur does not only have his own knowledge to dispose of, but also that of our 29 other entrepreneurs at Fortune Coffee.”

Machine learning enables you to discover patterns in large amounts of data. These patterns help to make optimal predictions or recommendations possible. In addition, the machine-learning model continues to improve itself, as it learns from previous predictions and the feedback on them. At Fortune Coffee, a model is built based on the order history of the past five years. By looking for patterns, customer groups have been formed with a similar history. The model was built with various Microsoft Azure tools. It was especially important to present the recommendations in the right place in the CRM system and make it easy to process them. Thanks to the entrepreneurs’ feedback, the algorithm that provides for the recommendations will only improve in the future.

Enthusiasm among entrepreneurs

Machine learning opens doors. Wiegant: “The system first learnt which products our customers buy. Based on that, a number of relevant products is proposed in combination with the products that similar customers have ordered. These products can directly be added to the order. A test with three entrepreneurs turned out to be successful, despite some start-up problems. After that, the system has been further improved, and now the solution is ready and presented to all Fortune Coffee entrepreneurs. Their response was mainly enthusiastic as they are happy with the support of the model in making relevant offers and improve the customer experience. Of course, there is also still some scepticism: does the model really know better than the entrepreneurs do with their customer contacts? Experience will tell.

Right now, it’s important that the entrepreneurs who are enthusiastic about machine learning will actually start using it. It’s easy. In the CRM system, only one extra block has been added where the recommendations based on history and other customers can be found. At the moment, five options are shown. It’s important that the entrepreneurs give feedback on the model. We used half a year for this. After all, the model must have the time to learn. Our customers already benefit from it as they get the most relevant offer. The entrepreneur now presents this from the solution.”

You don't need a large IT department

Wiegant continues: “Although we are a relatively small organization with no extensive IT department, we were able to implement this machine-learning project. Microsoft Azure ensures that smaller organizations can also benefit from major developments. Together with CRM Partners we have made complex technology practically available to our entrepreneurs. I look forward to their responses and can’t wait for the results we will have in half a year.”

Fortune Coffee continues to grow and with the use of new technologies it succeeds in building and maintaining personal customer relationships. Putting together all the information that entrepreneurs have in their heads in a machine-learning solution helps the company to better serve its customers. Wiegant: “I’m looking forward to the next innovative project together with CRM Partners and continue to realize an outstanding customer experience for our customers.”