Customer Churn: Why It Should Be Your Most Important Metric

Keeping an existing customer is much cheaper than acquiring a new customer. Here’s a guide to analyzing customer churn in order to improve retention.


“Leaky Bucket” Approach

Historically, companies have focused primarily on customer acquisition to solve customer retention problems.

Get a lot of people through the door and we are bound to keep some.”

“Customers are leaving, let’s spend on acquiring more customers.

This is what is referred to as the Leaky Bucket approach. The problem with focusing solely on customer acquisition is that you’re filling a bucket with a leak and hence, you’ll always be spending more and more on acquisition without ever filling that bucket.

Focusing on customer retention and expanding the revenues from your existing customers by up-selling them is a much more profitable strategy for growth than spending just on acquisition.

The cost of keeping an existing customer is at least 5 times cheaper than the cost of acquisition of a new customer. In addition, positive word-of-mouth from your existing customers leads to cheap, almost free customer acquisition.

Watch Your Churn Rate

In recent times, especially with the growth of subscription business models, churn is becoming one of the most important metrics that should be tracked by every business – be it a service like Netflix, or a SaaS company like Box.

The impact of even a 5% increase in churn can have significant effects on the revenues of a company. Consider a scenario where two companies have the same customer acquisition rate of 15%. Both companies start with 200 customers in 2015. Company A has 5% monthly churn and Company B has 10% monthly churn.

On the surface, a 5% difference doesn’t seem like a lot. But over the course of three years, these companies end at very different customer growth milestones simply because of the compounding effect of churn. Company A has 6,183 customers at the end of 2017, whereas Company B has 1,158.

That’s a 5.3X difference in customers due to a mere 5% difference in monthly churn!

Churn Rate

Analyzing Churn

Given that churn is such a critical measure of success for a company, it’s surprising how few companies understand why churn occurs.

One of the biggest mistakes I see companies make is that they assume that if they bucket their churn by demographics or plan type, it gives them enough insight into why churn happens.

For example, Company A finds that, when segmenting by plan type, 15% of their churn was in the starter plan, and 5% in the enterprise plan. From this observation, they draw the conclusion that there’s an issue with the starter plan that needs to be fixed.

But Company A’s customer base has 15% starter plans and 5% enterprise plans, hence those churn rates are as expected. They mirror the exact makeup of their customer base!

The problem with analyzing only churned customers is that this approach may not always expose the real causes of churn. Moreover, it doesn’t give you enough insight to make changes in your product or business strategy. For example, with the above insight, I can do nothing more than just know that 15% of customers churned are from the starter plan.

The other issue is that such ways of looking at churn are very simplistic and don’t fully reflect real world conditions. Customer churn can happen due to a lot of causes:

  • Product is not addressing the user’s needs.
  • Slow customer support.
  • Product cyclicality. People only use it during specific periods in the year.
  • Wrong channel strategy. Maybe you are acquiring customers from a channel that doesn’t fit your target customer profile.
  • Economic conditions. Every company is downsizing.
  • Pricing. Competitor is stealing your customers by giving discounts.

Know Your Customer 

The first step to understanding customer churn is to make sure you have the full context of your customer.

I cannot stress enough how important this factor is. You cannot analyze what you do not measure.

If you are a small company, you can call your customers to understand why they are leaving, but such approaches become unmanageable once a company scales. Additionally, customers may not always disclose the real reasons they are leaving.

Without data, you can do very little to understand churn. You have to capture not just the customer demographics, the plan type, and the price charged but also what marketing channel the customer came from, his/her usage metrics, support data, and email communications.

And, most importantly, you have to bring it all together into one consistent view.

Churned vs Active

Once you have the full view of your customer, it’s important to analyze not just who churned, but also who didn’t. Only then can you truly understand the causes of churn.

As I mentioned earlier, if 15% of your active customers are using the starter plan and 15% of your churned customers also use the starter plan, then the cause of your churn is not the plan type. However, if 30% of your starter plan customers churned, then yes, the starter plan has some issue.

You can do such analyses yourself by analyzing differences between two segments of customers — churned and active — across all the dimensions you track, including marketing channel, age, gender, business size, geography, usage metrics, support metrics, etc.

For example, from the table below, I can see that 30% of my active customers came from Facebook campaigns, but 60% of my churned customers came through that channel. Now I know from the data that my Facebook marketing strategy definitely needs attention.

Churned vs Active

When to Use Predictive Analytics 

As the number of signals you track increases, it can become cumbersome to analyze churn this way.

Bucketing across 50-100 different metrics can become not only time consuming, but lead to human errors as well. Additionally, not all differences among churned and active customers will be significant. That’s when predictive analytics can come to your rescue.

Statistical analysis and machine learning can help analyze churn at scale. So if you are at a stage where you have rich data and want to understand churn, you may want to consider predictive analytics.

The added benefit of predictive analytics is that it can even alert you when your future customers are likely to churn, so that you can proactively address the problem instead of reacting to it.

 Customer Delight

Customer Delight Is the Goal

To summarize it all, if you are a business chalking out your customer retention strategy, make sure you:

  • Collect and integrate all of your customer data into one place.
  • Don’t just look at churned customers, but analyze both churned and active customers to gain deeper insights.
  • And, most importantly, act on those insights to delight your customers!

For more information on customer churn and predictive analytics, check out PatternEQ. All images were provided by the author.

Ready to start your project?

Learn how ThinkApps can get your product launched faster, better, and with more value than you knew was possible.

@ThinkApps on Twitter