Customer Churn: Explanation, Reasons, Calculation & Anti-Churn Measures

In this post, we will deal with the topic of Customer Churn. Why is avoiding Customer Churn a primary business interest, what reasons are there for Churn and why is close monitoring a significant success factor? How can Churn be calculated and demonstrated? What are the countermeasures available to fight Churn? We answer all these questions in this post. Based on years of experience in churn management in different industries and companies, we provide recommendations at the end of each chapter and want to give you valuable tips & tricks that should make understanding, measuring and managing Churn easier.

Churn - Definition of the term and manifestations depending on the type of customer loyalty

The term Churn has now entered international business language. It comes from English and means "migration". As a rule, and when it comes to the term Customer Churn, even more so, it is about customer migration. Depending on the business model of the company and the nature of customer loyalty, there are different manifestations of Churn.

Customer-initiated Churn (active Churn)

In the so-called active Churn, customers actively make the decision, to end the customer relationship with the company. The type of customer loyalty determines what they have to do to end the customer relationship.
If it is a voluntary form of customer loyalty (e.g. hairdresser, cinema, restaurant, supermarket) customers simply stop shopping. This can happen without warning for the company, as it does not require any form of communication or cancellation by the customer.
If, on the other hand, it is a form of contractual customer loyalty (e.g. telecommunications contracts, contracts with electricity providers or insurers), written cancellation is usually required to cancel the existing contractual relationship. This cancellation should be understood by the affected company as a very serious warning that those who cancel want to end the customer relationship. You, as the person in charge of customer loyalty, should become active now and initiate appropriate customer loyalty measures - best oriented towards customer value. More about this later. By using statistical prediction models (Predictive Models), it is also possible to initiate customer loyalty measures preventively, before receiving the cancellation. Because these models master the calculation of a churn probability (Churn Propensity/Churn Risk) of customers. We will explain why the use of such models is useful later in this article.

Company-initiated Churn (passive Churn)

In addition to the active Churn explained above, there is also the so-called passive Churn. "Passive" because it does not come from customers, but from the company. Examples of passive Churn are the deactivation of mobile phone contracts due to non-payment by the customer or the cancellation of insurance contracts by the insurer (e.g. because the customer has become unprofitable due to high claims amounts).

Why every form of Churn is critical and to be avoided from a company perspective

Customers represent sales or at least sales potential. Therefore, all companies are interested in building and maintaining as broad a customer base as possible. In addition, companies invest in their new customers depending on the industry and business model. Furthermore, long-term customers usually prove to be more profitable (Reichheld 1997, p. 54 & 58ff.; Rust et al. 2000, p. 96). Therefore, it is necessary to tie attractive customers to your own companyand at the same time continuously enlarge the customer base by acquiring attractive customers.
A major advantage of investing in the loyalty of existing customers is that these existing customers are familiar with the value of the company. Thus, all customer loyalty activities can be oriented towards long-term margin. New customers are unknown per se and there is a risk of overinvestment, which cannot be compensated for a longer period of time by positive contribution margins from the customer relationship.

Reasons for Churn

In the further course of the article, we focus on active Churn. Reasons for active Churn can be very diverse. First, there are personal reasons. For example, the personal circumstances of the customers may have changed in such a way that the offer is no longer relevant or interesting. Competition is also a major reason for active Churn. More attractive offers from another provider can lead customers to end the existing relationship with a company. In addition, price, service, and quality usually play an important role among the reasons for Churn.

Reasons for Churn versus Reasons for Cancellation

At this point, we also need to differentiate between reasons for Churn and reasons for cancellation. Despite considerable overlaps, there is a major difference in when and how the company learns about the dissolution of the customer relationship. If there is a contractual customer relationship, customers usually have to cancel the contractual relationship in writing, as described above, in order to be able to churn. The reason for cancellation is usually mentioned in this cancellation. The company should carefully record this in the customer's account in order to make customer loyalty offers based on it and on the basis of customer value. The company usually does not learn about the reasons for churn, or only when the event Churn is a thing of the past, i.e. usually when there was a voluntary customer relationship, in which no (written) cancellation was required. (Example: A customer stops visiting a restaurant at a certain point in time. The restaurant owner will hardly find out why that is.)

Relevance of the Reasons for Cancellation/Churn for the Company

As written above, the first goal for those responsible for customer loyalty is to bind the company's customers as successfully as possible to the company. This bonding can only succeed if the company closely monitors the reasons leading to cancellation or Churn, takes them seriously, and tries cross-functionally to remedy them. Thus, the service experience for customers must be continuously optimized, the price critically questioned, the quality increased, and work must be done on the attractiveness of our own offer compared to the competition, if this emerges from the reasons for cancellation/Churn. Therefore, it is recommended to conduct in-depth Customer Churn surveys at regular intervals in order to learn more about the main reasons for cancellation or Churn of your own customers.

How do you calculate Churn?

As described above, it is recommended to consider both the receipt of cancellation and Churn as KPIs. Both as an absolute number and as a quota.

Absolute number of cancellations received

This is the quantity of absolute cancellations received. Usually, this number is shown on a monthly basis. To better understand and manage this key figure, it is recommended to break down the total number into meaningful categories. For example, by product, region, sales channel, and customer segmentation. This granular presentation reveals trends over time that need to be carefully observed and actively managed.

Quota of incoming cancellations on the customer base

To calculate the quota of cancellations on the customer base, the absolute number of cancellations of a month is divided by the total customer base and the result is multiplied by 12. We recommend reporting this or another quota in addition to the absolute number of incoming cancellations. The reason is simple: If there are noticeable changes in the size of the customer base over time, this will probably also be accompanied by an increase in the absolute number of incoming cancellations. A focus on the absolute number can cause alarm. However, if you divide by the also growing customer base, the picture is put into perspective.

Quota of incoming cancellations on the potential for cancellation

The quota of incoming cancellations on the potential for cancellation is calculated similarly to the quota of incoming cancellations on the customer base described above. However, the basis is not the total customer base, but only those customers who could currently still cancel - the uncanceled customer base.

Recommendation

We advise on the presentation of the absolute number of cancellations and the addition of one of the two quotas. Statistically cleaner is the use of the quota of incoming cancellations on the potential for cancellation. Because in particularly strong acquisition months, the quota on the stock will be diluted by the increased stock. Conversely, it is artificially raised if customer acquisition does not perform in the respective month.

Absolute Churn

This number represents the real customer losses recorded in the respective month. So it is the customers who have actually been lost after the receipt of the cancellation and unsuccessful Churn measures. In telecommunications, for example, these are deactivated SIM cards. For insurance companies, cancelled contracts. At the hair salon, it's those who have decided to visit another hairdresser in the respective month. As with the number of incoming cancellations, it is recommended to break down the number into granular categories to quickly and early recognize trends.
It is recommended to break down this total number into active and passive Churn.

Churn rate on the customer base

In the Churn rate on the customer base, the absolute active Churn of the respective month is divided by the total customer base and the result is multiplied by 12. Like the corresponding quota of incoming cancellations, this quota has weaknesses, as it can be heavily influenced by effects in the customer base.

Churn rate on the Churn potential

To calculate the Churn rate on the Churn potential, a certain preparation is required. The Churn potential has to be calculated continuously. The Churn potential includes customers who, based on their contract term, would have had the opportunity to end their customer relationship in the respective month of consideration. Example: A customer signs a contract in March 2023 with a term of 24 months and thus belongs to the so-called Churn potential in March 2025. Even if the customer should extend his contract during this time, he remains part of the original Churn potential in March 2025. If he extends again for another 2 years in May 2024, he also appears in the Churn potential for May 2026. This is a cohort consideration. All passive deactivations are subtracted from the Churn potential. To calculate the quota, only the absolute active Churn of the respective month is divided by the Churn potential. The subsequent multiplication by 12 is omitted. The reason for this is that the Churn potential represents a monthly value, whereas the customer base is considerably larger and should be understood as an annual value. Passive churners are also deducted from the Churn potential.

Recommendation

A quota calculated on the customer base, as described above, is subject to a considerable bias due to effects on the customer base. For this reason, we strongly recommend presenting Churn as an absolute number, supplemented by the Churn rate on the Churn potential. A transparent presentation of all cancellation and Churn KPIs in the respective granular details by product, region and sales channel in a transparent Customer Churn Dashboard helps all participants to quickly draw the right conclusions and initiate effective countermeasures.

How do you design effective countermeasures? 

As described above, the company has two critical events in the customer lifecycle to avoid: The receipt of cancellation and/or Churn. Two different types of measures can help here: preventive and reactive.

Preventive customer loyalty measures 

Derived from the word, this type of measure is designed to prevent incoming cancellation and Churn from the outset. These can be divided once again into triggered and untriggered measures.
Untriggered measures are measures that you implement throughout the entire customer lifecycle and that help to exceed the expectations of the customers. This exceeding of expectations creates customer satisfaction, which leads to higher customer loyalty and ultimately to increased customer retention (based on Bruhn/Homburg 2000, p. 10).

The more successful you have been with these measures during the customer life cycle, the less likely customers will receive a cancellation/Churn.
Nevertheless, you will not succeed with all customers, as there will inevitably be customers whose life cycles will be followed by the crisis phase and then the separation phase after the maturity phase (Dittrich 2000, p. 131).

At this point, it is important to use trigger models and triggered measures during the crisis phase to address the customers as early and as effectively as possible. Thus, a crisis can become a new beginning and customer migration can be prevented.

Trigger models for the early detection of cancellation receipt and Churn (Predictive Models)

The goal of these models is to predict the probability of receipt of cancellation or Churn (Churn Propensity/ Churn Risk) on an individual customer level (Customer Churn Prediction). To do this, a score for the respective event "receipt of cancellation/Churn" is calculated on an individual customer level. Usually, a linear regression is used to create such a statistical model. This learns from the customer behavior of the past. As always with statistical surveys, it is recommended to have a solid database. What you need, therefore, are sufficient data from customers who have cancelled/Churned in the past. In addition, you need the behavior data in the months before the event "receipt of cancellation/Churn". The linear regression looks for similarities in the behavior of those who cancel/Churn in the weeks/months before the respective event. These similarities must statistically significantly differ from the behavior of those who have not cancelled/Churned. If such similarities have been identified, it can be tested whether these were only a temporary phenomenon and only applied in a certain time window considered in the analysis or whether those who cancelled/Churned in the last weeks/months before could also have been recognized by these similarities. If this is the case, parameters with discriminatory power have been identified that can now be used to look for customers in the present who show the critical behaviors in a certain month relevant to the analysis in the customer life cycle. If these customers can be found in the customer base, an immediate selection for contact with a campaign relevant and attractive on an individual customer level is advisable, with the aim of preventing the cancellation/Churn preventively.

Reactive customer loyalty measures

These measures are triggered measures. They occur in response to an expression of will by the customer to end the customer relationship. This can be a cancellation if there is an existing contractual relationship. Without an existing contractual relationship, however, this can also simply be an expression by the customer to choose another provider of the respective products or services in the future. Similar to the preventive measures, the high Churn score serves as a trigger here. A cancellation or intention to cancel or the occurred Churn serve as a trigger. An example of a measure after apparent cancellation without a contractual relationship are coupons from the provider FREE NOW. When the taxi service from FREE NOW has not been used for a long time, the customer receives a push message in the sense of "How about us again - 30% off your next ride." Here, the fact that the customers have not used the service for a long time serves as a trigger for campaign selection. The value of the customers when they were still using the service determines the discount at this point.

Recommendation

Now you can rightly ask the question, which is better - preventive or reactive customer loyalty. Our recommendation is: Use both options when the opportunity arises. Because not every company has customer data available or generally the possibilities to use scoring models. It might be difficult for hairdressers, for example. Both approaches have their weaknesses. Predictive Models may be brilliantly developed. Yet, they never predict the future with a validity of 100%. There are scatter losses. Customers with a high score do not cancel and customers with a low score cancel anyway. The advantage of preventive contact is undeniably that the customers have not yet or not yet long made the decision to turn their back on the company. The contact still occurs quite early in the crisis or separation phase. Therefore, relatively manageable investments in customer loyalty can compensate for the scatter losses with clever control at the end. The reactive measures do not have these scatter losses. Because the company only reacts when the customers have already acted and have clearly expressed their already made decision to separate. For this, the level of required investments here is considerably higher because an already made decision needs to be revised. If your goal is to bind as many customers as possible successfully to your company and if the business model and data situation allow the preventive approach, we therefore recommend going ahead both preventively and reactively.

What tools can be used to implement these countermeasures effectively?

Customers want to be individually approached with offers relevant to them, in the channel relevant to them. Thus, you will do yourself a favor if you use some form of segmented approach. Because the more appealing and relevant the communication measure and the customer retention offer are for the customer, the higher the conversion rate will be.

Depending on the size of your customer base, the number of resulting segments can be high. From the factors product, customer segment, possibly age segment, customer value, preferred channel and possibly others, a large number of communication measures result. To implement, send and subsequently evaluate all of these, the use of a CRM tool is recommended. There are a multitude of providers and products on the market.

Examples of tools from the CRM category include:

SugarCRM

Salesforce CRM

BSI Customer Suite

monday sales CRM

SAP CRM

Zendesk Sell

HubSpot CRM

Pipedrive

Microsoft Dynamics 365 Sales

Which of these and other tools is ideal for you and your purpose needs to be defined based on a requirement analysis.

Conclusion 

No matter what size your company is. No matter what industry you are in. Doesn't matter if it's B2B or B2C. Effective customer retention is essential for business success. Through the right, ideally customer-specific, approach with an attractive customer loyalty offer at the right time, you can extend your customer relationships successfully and thus make a decisive contribution to the success of your company. The most important findings for you are:

• The more you preventively reduce the Churn with effective measures, the easier you can dissuade the customers from the sometimes not yet made decision to end the customer relationship.

• It is crucial to deal in detail with the reasons for cancellation or Churn of your customers. Only if you know these reasons, effective countermeasures can be initiated.

• Not all customers are equal - therefore it helps you to break down cancellations and Churn into granular levels in order to better understand and subsequently combat trends.

• Present these KPIs transparent for you and other stakeholders in a Churn-Dashboard.

• Choose the KPIs for your Churn Dashboard that make most sense for you and your company and that can be calculated with your possibilities.

• Choose a customer-centered approach for the design and implementation of countermeasures. What could your company do to bind you successfully if you had the respective reason for cancellation or Churn? And then test and learn how your customers react to your offers.