Customer Churn: Why It Happens & How to Reduce It?

Jan 30, 2020

11 mins read

Tanuj Diwan

You might have heard or read about this term called “customer churn” but do you know what does that really mean? Let us take you through on what customer churn is and why it is one of the most important aspects of evaluation for a growing company.

Customer churn is a reality that spares no one, whether it is the largest, most successful companies or a small startup.

It is of utmost importance that companies Evaluate and Understand why those certain percentages of customers abandoned your ship and find a new solution instead of yours.

what exactly is customer churn and why is it important?

Customer churn to simply put is the percentage of your customers who stop doing business or end their relationship with your company within a given time period. To even simply further, it is any business you lose.

For example, if you take a subscription-based business like Salesforce, customer churn is of a top priority since the Customer Acquisition Cost (CAC) is higher than to a cross-sell or upsell to an existing customer.

Another example can be in case of ad agencies which only have a few large clients and if one of them churns, it will become a big issue for them. CAC for a new customer can be as much as 5 times that to retain a customer.

Research says that there is More than a 25% increase in profit if a company increases customer retention by 5%.

Let this sink in for a moment. With that statistics in mind, wouldn’t you want to put your efforts and strategy in retaining your existing customers by nurturing them

and keeping track of the customer retention and churn rate? This brings us to our next point on calculating the customer churn rate.

How do you calculate the customer churn rate?

Calculate Customer Churn

Photo by Roman Mager on Unsplash


To put it into the calculation, you would need to figure out a time period and the number of customers who left during that time period, divided by the number of customers at the beginning of that time period, multiplied by 100%.

How to Calculate Customer Churn, Customer Churn Calculating Formula,

So, with keeping in mind this formula, let’s say you started the month with 100 customers (to make it simpler) but lost 5. Then using the above formula, this would bring your churn rate to 5%. Simple right?

As simple as this might seem to be, being able to calculate and keep track of the churn rate would be important to assess how your business,

campaigns, various projects, etc, are performing. While the actual result (5%) might not seem to show much but it Lets you dig deeper into the root cause and make appropriate changes.

 how do you predict customer churn?

Data, Data, and more data. That’s right. Data is the basic layer to predict customer churn.

Analyzing the past customer churn data and assessing the behavior of the customers before they churned will help you predict and stop it from happening by making the necessary changes.

How to predict customer churn

Photo by Jen Theodore on Unsplash


Let us split this into three steps –

  1. Gather past customer data.
  2. Upload data onto a prediction service.
  3. Use the analysis of each customer to see the risk of churn.

Gather past customer data

Machine learning is the key to predict customer church. This is done by artificial intelligence (AI) where analysis is built based on the comparison of data examples.

But to make these comparisons, we need to identify customers who were impacted in any way when they churned and when they did not.

There are 4 types to segregate different features of a customer –

  • Customer features detail about the customer like name, gender, age, etc.
  • Usage features: customer’s interaction with your product or service.
  • Contextual features: any information that might assist you with the analysis.
  • Support features customer’s interaction with your support team.

Feature segmenting might change depending on the type of business you might have, especially for usage and contextual, since they are more specific to the product and business.

Once you determine these data, you would need to decide on the time frames. This can largely vary based on the type of business.

For example, for a subscription-based business that might provide yearly billing, one would take the previous year into the time frame consideration to assess the four features.

Similarly, for most of the companies out there, you would take 2 to 3 of the previous months to evaluate customer information and features to calculate the fact if the customer would churn or not.

Once this is done, you need to extract the data across the time frame and arrange it on a spreadsheet for the next step.

Upload data onto a prediction service

This is the step that we love. Easy and simple. With prediction services such as Google Cloud ML Engine and BigML, life has become much easier.

Our personal favorite is BigML. You can go ahead and create a free account in BigML (Now who doesn’t like free stuff!) and upload your data on the spreadsheet.

It will take you to step by step and their user interface is pretty friendly, so fret not.

Once you upload your data and go through the steps, it will bring you to a model pane.
Customer Churn Analysis

Once you get here and click on the dataset, BigML will take you to an analysis called decision tree model which looks something like this –

Define Customer Churn

Yes, it might look overwhelming and might not look like a “tree” but let us make it simpler for you to understand.

So, the way to read this is to know that the nodes are associated with a feature value from your data input with several answers which are represented by the branches.

You start from the root node and depending on the answer you choose it will take to the eventual “leaf” or say the answer for prediction.

Time to make those predictions

So, what have we done until now? Created the spreadsheet with customer data from X months ago, uploaded it onto BigML, and went through the model pane and decision tree model.

As mentioned earlier that we need a comparison between the customers X months ago and customers now, we need to repeat the same steps but with the time period to be now and not X months ago.

There is only one change this time around where the database doesn’t hold the churn details, so you need to create two separate columns with churn predictions in one and confidence levels for all these predictions in the other.

After you run through this data in BigML, voila you will get these predictions to take appropriate actions to solve it.

Companies like Groupon use this strategy to predict their customer churn which leads us to strategies to reduce customer churn rate.

How do you reduce customer churn rate?

How to Reduce Customer Churn

Photo by Markus Spiske on Unsplash

These are some of the aspects that businesses need to evaluate to reduce the possibility of customer churn.

Almost 70% of the reason a customer churn, has nothing to do with the product. It always comes to the quality of customer service, which is paramount for any business.

As long as you solve for your customers with genuineness, you should be on the right track. Let’s get right into it!

Top-notch customer service

Discover Bank has one of the well-trained customer service representatives and user-friendly systems in place to keep their customers satisfied in every step of their communication with the bank.

Customer Churn Rate

Making that first good impression whether it is through the frontline sales representatives or the customer support team, first friendly interaction helps prospects to easily fall in love with the brand.

Customers need to feel that they are being taken care of in a genuine manner and the company is there to solve for them. 91% of unhappy customers don’t want to do any business with you again.

Customer service is of key importance to keep the churn rate low.

Put efforts to retain the best and most profitable customers

60% – 70% is the probability of upselling and reselling to a satisfied customer while 5% – 10% to a new prospect.

That should say it all. The most traditional way to go would be to incentivize or put your efforts in retaining a customer who is about to churn but the question you should ask, is this going to be profitable to the company.

The stats say otherwise. Since a happy customer is already satisfied and trusts your brand brings them way down the funnel rather than starting from the top.

Build your competitive advantage

HubSpot and ThriveHive might offer similar products but most of the people are not aware of ThriveHive. That’s because of the thought leadership content and free academy course HubSpot offers.

It shows them as a leader in their industry which in turn helps them earn trust by setting them apart.

With the free content HubSpot offers, it brings them in front of the customer’s mind and increases word of mouth. Build that competitive advantage to contribute to higher profit, loyalty and increase customers.

Customer Churn

Communication is key

Communication is, has and will always be one of the most important aspects to run a successful business with satisfied customers.

Being proactive with your communications with your customers, right place and at the right time is the formula businesses to evaluate and invest their efforts on.

Here is one of the emails I received from Audible to sign up for the Audible Challenge 2020 which is to the point, clear and concise.

Whether it is social media, emails, calls or chats, whichever channels that your customers use, your business should be proactive in reaching out and also addressing their queries in no time.

churn analysis case study "Audible"

Create a community around your offerings

People look for a feeling of belongingness where likeminded people can gather and communicate with each other to walk this journey.

Creating a community around your products and services helps customers to feel that they are part of your brand and lowers the risk of churn.

Moz is a great example where they have built a vibrant and tight community around its services. Within their community, they run a guest blog,

where anyone from the community can submit a blog to be published. They keep their community active and engaged with their Q&A, results, queries, etc creating the front of mind for their brand.

churn analysis case study "Moz"

Make it very difficult for customers to ignore your brand

Businesses to strategies to hook your customers up with your brand in a way that makes it difficult for them to leave and ignore you or switch to your competitor’s product or services.

Zapier is a great example where it gives the prospects a 14-day trial of their premium subscription with the same customer support, tools and features that a regular professional customer would have access to.

At the end of the 14-day trial, the prospect would have experienced the possibilities of the tools and seen the results which would make it tough to leave the product.

Feedback is the breakfast of champions

Yes, it sure is. Business needs to provide an easy way for customers to provide regular feedback for them to evaluate and solve for the customers. But that doesn’t mean you send out a feedback request email to your customers just for the sake of it with saying something like

“We will use this to improve your experience”.

Although is basic and you would want to be more specific and give them ownership.

UnSplash where most of the pictures in this article are from, does a great job with Customer feedback. It’s specific when they say that this will help them prioritize what to build next while giving the customer a sense of ownership of getting a better chance to get their favorite featured developed.

churn analysis case study "UnSplash"

Wrapping things up!

There you have it, folks. All you need to know about customer churn. Businesses need to evaluate the strategies outlined in this article and you will be well on your way to lowering your customer churn rate and increasing your profitability. Churn down!

How much did you enjoy this article?