Skip to content

Aarstiderne gains unique insight into customer loyalty

Case: Aarstiderne

twoday's Customer Churn Framework identifies customers who are at risk of becoming inactive

case-aarstiderne-hero

twoday's Churn Framework makes predictions and tells Aarstiderne which customers are at risk of churning – with great precision.

In the past year, Aarstiderne has carried out a strategic customer loyalty project entitled “the best customer experience each time”. The objective was to give more structure to the customer loyalty work. What did it take to achieve this goal?

“We made a list of various initiatives, one of which was to gain insight into which customers are most prone to churning,” says Carsten Dreyer Christensen, Head of Business Analysis.

After discussing with twoday, Aarstiderne chose to participate as a pilot customer in the twoday Churn & Retention Framework, which, based on a specific Advanced Analytics model and Machine Learning, is able to identify customers at risk of churning.

“It sounded like a win-win for both parties,” says Carsten Dreyer Christensen, continuing: “twoday needed a customer with relevant data for the project, and we needed a tool to help us further improve customer loyalty.”

Before initiating the development work for the actual model, clear objectives were defined, including what the model was supposed to do, the model change options level, and how simply it could be done.

Customer Churn Framework

The first edition of the model was based on 25 different parameters, including transaction data, marketing definitions, web and app activities, customer supply chain experience as well as basic master data. All of these data were fed into the model – and already at the first round, the model was able to differentiate the customers.

Over time, the model has been further developed and refined, now working with more than 300 different variables. Concurrently with this improvement, the precision has increased significantly.

Within a few months, when the model’s precision has become even better, Aarstiderne will use the results to increase the customer experience for real.

"The model now tells us with great precision which customers are at risk of churning as well as which customers are deemed loyal. This will not save the world, of course, but the solution’s value potential is very important!"

Carsten Dreyer Christensen, Head of Business Analysis, Aarstiderne

Being at the centre of a pilot project

As mentioned earlier, Aarstiderne’s solution was delivered as a pilot project. In this respect, Carsten Dreyer Christensen says:

“It has been a really great and interesting phase that has been instructive for both parties. twoday has invested a lot of energy in the project, resulting in a well-functioning solution that gives us great value.”

Carsten Dreyer Christensen also emphasizes the high amount and good quality of Aarstiderne’s data as an important factor for the project, thereby forming a good basis for the churn & retention solution.

About being a first mover, Carsten Dreyer Christensen says: “It is interesting that Aarstiderne is among the first companies to use Machine Learning. I find that awesome, actually.”

 

About Aarstiderne

Aarstiderne produces and sells organic food. Aarstiderne is a limited company founded in 1999 by farmer Thomas Harttung and chef Søren Ejlersen.

Aarstiderne employs more than 150 employees in Denmark and Sweden and had a turnover in 2016 of DKK 566 million (USD 80 million) and a profit after tax of DKK 38 million (USD 5.4 million).

Approx. 60,000 customers in Denmark and Sweden enjoy fresh food from Aarstiderne.

case-aarstiderne-farm