Yet despite this, many brands today will still focus a larger effort on customer acquisition than retention, often to their detriment, so why does this happen?

“Attracting new clients requires more resources, time, effort and this is why so much focus is put on it. For many brands the experience that consumers get once converted to customers may actually be worse in some ways. For example, marketing will produce a series of bespoke emails to convert a prospect into a client but once converted they may only get a monthly newsletter. Retention is often not treated with the level of importance it should.” Paul Carney, Managing Director

Today’s consumer is more informed, more empowered and can more easily change provider. As such it is critical that brands leverage their first-party data and put a clear emphasis on the retention of customers. You need to understand the motivations behind customers leaving and be able to forecast the impact that different measures would have in reducing this.

A Churn Rate (also known as the rate of attrition) is the percentage of customers who stop using your product/service. Understanding this can help you to:

• Calculate and improve Customer Lifetime Value
• Measure your company’s health and long-term prospects
• Identify changes that had an adverse effect on retention
• Forecasting brand /service/segment performance
• Understand the motivations and key intervention points where customers leave

Yet understanding your Churn Rate is no simple task…

The surprising complexity of calculating Churn Rate

It is said that “people vote with their feet” and when a customer stops using your product and service it’s rare that they will give you a fair warning or detailed explanation. As such understanding churn is critical for any business, not only so they can identify factors that correlate to customers leaving but so that they can proactively predict ‘at-risk’ customers and prevent churn.

In its most simple form, churn can be understood as:

Customers lost during a period ÷ Customers at the start of a period

While this equation is easy to wrap our heads around, the way in which we define these different aspects can drastically change the outcome. Here are the factors which make calculating churn surprisingly difficult:

Defining the moment of churn

While most businesses understand churn as the point in which a customer stops doing business to them, in most cases, the customer’s motivation and timing is unclear.

For example: A supermarket brand may have loyalty card data that customer X visits every week to do their shopping. If one week they don’t visit the store, does that mean they are a churned customer? Could it be they just forgot their loyalty card? Could it be they were away? The truth is you won’t be sure until much later and will have little data as to why that customer stopped shopping with you.

One type of business that might have a clearer view on churn are those with a subscription model, but even this has its complexities. A subscription business might define churn as the moment a subscription ends and a renewal doesn’t happen, or the moment of cancellation. Yet, if someone cancels but their subscription hasn’t ended, you still have time to win them back…so should you include this in your churn rate? Should you include people on a free trial/new sign ups in your churn rate? It won’t be possible for these customers to churn in the time period of their trial/first month of payment and so could distort your calculations.

As we can see, understanding the moment of churn is not as straight forward as is desirable. Churn is typically a lagging indicator that can’t be measured in real-time and can easily be distorted.

Defining your amount of customers

How many customers do you have? It seems like a really simple question doesn’t it, and yet its one of the problematic areas you’ll need to address when thinking about Churn Rate. This is because the value of churned customers is affected by the entire period of time, where the number of customers is affected by the beginning of the period of time.

Let’s say you’re trying to figure out a monthly churn rate:

On day 1 you have 100 customers.

You lose 5 of these customers throughout the course of the month.

By the end of the month you’ve gained 400 customers.

You would have two very different rates of churn depending if you looked at amount of customers at the start of that period (5% in this example) or end. (1%) No matter which number you use for the ‘total number of customers’ it will be pretty distant from the day 1 number, the day 30 number, or both.

The total number of customers is also highly affected by the type of customer they are. Whilst it is generally true that newer customers are more likely to churn, new customers may be on a trial or have pre-paid for their first month of services and therefore may not be able to. You also have to consider that if a large number of new customers come in at the end of the month (quite a common scenario when we think about how payday affects purchases) they can potentially disrupt the sample size and cause mis-leading data.

Sample sizes are not linear

We are all familiar with the concept that it is easier to maintain 10 relationships than 100. For fast growing companies this means that your Churn Rate can fluctuate wildly month-on-month as you acquire more customers. If you had 1000 customers and a Churn Rate of 10% then you might forecast a similar rate as you grow, but you may not have the same infrastructure to support a larger amount of customers. You can imagine the impact on your business strategy if your rate of churn grows significantly overnight! This highlights that whilst Churn Rate can be a useful metric, it is still limited in its scope.

Benchmarking is difficult

Whilst getting a grip of your own Churn Rate is useful, how do you know if it’s any good? How does it compare to your competition? This desirable information is often difficult to come by and while there may be some industry-level stats you can find in the public arena, this can only be used for broad comparison. For example, the average Churn Rate for app users is about 80% within 90 days…but what kind of apps, in what location? A gaming app might be set up to encourage repeat visits where a travel app might only be used in specific situations. None of which is particularly helpful to you if launching an app today!

Not all customers churn equally

One of the biggest problems with a static Churn Rate formula is that it will assume that a customer is equally likely to leave at any time. This is something we know isn’t true. Typically newer customers are more likely to churn and as time goes on the rate of churn will decrease, but never disappear. Most businesses will recognise that there is a natural cycle to customer churn. Some customers, despite never having fault with a business will simply try a new provider after a period of time. Almost the business equivalent of a “12 year itch.”

As such when we think about churn in relationship to the Customer Lifecycle we should picture a curved graph. This can help us to identify key points in time to increase our engagement with customers.

As well as considering how customers at different lifecycle stages will churn at different rates, we must also factor in how segmentation will affect Churn Rate. Different segments will all have varied buying styles and relationships to specific offerings.

For example: You might have a ‘full’ offering and a ‘light’ offering. The light offering may attract a higher amount of new customers, but may also have a higher Churn Rate than the full offering. One way to interpret that data would be that you should focus more attention on getting customers onto your full offering due to the increased likelihood they will become long-term customers….but this may be a false reading. The reason that fewer full offering customers leave may be that they have already experienced your brand/service in your light offering and then upgraded. If you recruit more people into the full offering you may actually find that the Churn Rate is higher.