Reducing SaaS Customer Churn with Predictive Analytics

Why to use Predictive Analytics?

Predictive models can serve a variety of goals. The objective of this blog post is to investigate how predictive analytics can be used to reduce Customer Churn in the SaaS industry.


The biggest shift has been in the power of SaaS customers to influence their buying experience. With access to global products and services, they can demand better quality, lower prices, and faster delivery and the only way to survive in this new economy is to embrace and leverage the power of information.

A survey conducted by Pacific Crest, showed that over 30% of SaaS providers have an unacceptable level of churn, so, even if a company have a big volume of new customers, is important to look at the number customers that are leaving from the back door as this may be affecting company growth more that you think.

Monthly Churn and Annual Churn Rates

In general the Monthly Churn Rate is calculated by subtracting the number of customers remaining at the end of a month (Cend) from the number of customers at the beginning of a month (Cbeging) and divided by the number of customers at the beginning of the month (Cbegin).

Monthly Churn Rate               =             (Cbegin – Cend) / Cbegin

The Annual Churn Rate, is then calculated by multiplying the monthly churn rate by twelve.

Annual Churn Rate                 =            Monthly Churn Rate * 12

Note: New customer acquisition should always be excluded from the number of customers remaining at the end of the month.

In the following table, you will see how even a 2% Monthly Churn Rate became a staggering 21.53% Annual Churn Rate!

I just use the formula: Annual Churn = 1 — ( 1 — monthly churn )^12.

Monthly Churn  Annual Churn
0.50% 5.84%
1.00% 11.36%
2.00% 21.53%
3.20% 32.31%
4.00% 38.73%
5.10% 46.64%
6.00% 54.41%
7.00% 58.14%
8.00% 63.23%

Ok, now, we understand how critical is to tackle the Churn Rate, let’s now have a look at the data we need to collect to build a Churn Model.

Churn Model for SaaS Providers

Every business is different, and your customers behave differently from business to business. My hope is not that you apply the exact same model, but that you apply these principles and customise your own. This is only the initial stage to apply Predictive Analytics, it will take time, but, it will worth it.

The most common churn prediction models are based on statistical and data-mining methods, such as logistic regression and other binary modeling techniques. These approaches offer some value and can identify a certain percentage of at-risk customers. A more accurate approach to customer churn prediction is the method of calculating customer lifetime value (LTV) for each and every customer.
Customer Life Cicle Touchpoints

By merging exact micro-segmentation of this touch-points with a deep understanding of how customers move from one micro-segment to another over time – including the ability to predict those moves before they occur – an unprecedented degree of accuracy in customer churn prediction is attainable.

Once those customers at risk of churning have been identified, the company is now empowered to use targeted proactive retention strategies  to run on each individual customer to maximize the chances that the customer will remain a customer, but this a topic for another blog post!

Useful Links:

Leave a Reply

Your email address will not be published. Required fields are marked *