Arcum

Arcum Partners began when two former colleagues working on helping newspapers reduce subscriber churn stumbled upon the payments industry. Both Sebastian and Tad had worked with tremendous amounts of data but had never experienced the rich quality of payments data. By working with some of the largest organizations in the payments world, they became deeply knowledgeable in all aspects of the payments process, including their problems.

Shortly after landing their first pilot client, they reached out to Mikhail Dmitriev, Ph.D., an assistant professor of macroeconomics at Florida State University to come in as an advisor. Mike's advisory role quickly expanded, and just a couple of months later, he was helping the team code the first retention algorithm.

A short year later after the team got together, they analyzed over 150k unique merchants across five different payment companies and created an algorithm capable of identifying clients at risk of leaving up to 7 months in advance with 80% accuracy.

The Problem: Reactive Retention

One of the biggest problems we identified in the payments industry was customer attrition. in fact, 1 out of every 5 clients is expected to leave their payments provider every year (Goldman Sachs). In terms of sales volume, attrition costs the industry, roughly $1 out of every $10 processed (TSG). This problem is largely due to the fact that the industry still relies on reactive retention campaigns.

Problems with Reactive Retention:

1) Difficult to know when a client is unhappy

2) Cost the company money as a result of inefficiencies

3) Customers receive "discounts" or "perks" only after they have decided to leave

Market Size: A $7 billion annual problem

According to Statista, there are roughly $7 trillion dollars in sales volume from credit cards every year. Since 10% of all the processing volume is expected to attrite annually, this means that about $700 billion in sales volume is lost from customers leaving. In terms of revenue, payments companies make about 1% in net revenue of all sales volume they process, meaning that about $7 billion is lost every year as a direct result of customer churn.

Solution: Proactive Retention

In order to solve this problem, we developed a machine learning model capable of identifying customers that will leave and the reasons why. That way our clients can develop proactive retention strategies for their customers.

In our most recent case study, we identified 40 accounts in one month out of a 5k merchant portfolio. 7 months later, 80% of the customers identified by our model had left our client. By identifying customers that will leave in the future and the reasons why we provide the necessary tools for our clients to develop proactive retention campaigns that actually work.