Over the last 4 years we have been on the quest to build a SaaS Churn predictor, with building pieces over time and understanding the moving parts
In this article we are going to show how to build one for your organization and we have divided the sections into multiple parts, you can directly go to a part that suits you
What is churn?
Why do you need one?
How to build a SaaS churn predictor?
Optiblack SaaS Retention Program
Resources under the program
Video overview on how to run it
The Software as a Service (SaaS) industry is one of the fastest growing industries in the world.The SaaS companies are no doubt getting well equipped with all the growth parameters they need. They are getting specialized on each vertical that contribute to the overall growth of their businesses. But with such an immense growth comes equally challenging hurdles as well that companies need to overcome. And one such major challenge which is a common plight for almost every SaaS business is Customer Churn. So far the industry is still in its infancy and there is no standardized approach in dealing with this topic. All organizations are exploring, learning and growing in their own ways. Most probably that is the reason that has brought you here to explore about Customer Churn.
What is customer churn?
Customer churn, or customer attrition as it is sometimes referred to, is an event when a customer stops using your product resulting in their non-renewal of the subscription. In other words, it is the end of the relationship between you and your customer.There could be multiple reasons for them to stop buying your service. They can vary from the dissatisfaction from your service to their inability to derive value out of it or any other reason. Instead of digging more into why the customers leave your company, let us spend some time on understanding the impact customer churn has on your business.
Impact of customer churn on business
If you are wondering what impact customer churn creates on your business then you have to take into account the
, just a 5% increase in customer retention rates could potentially boost profit up to 95%!
If you will look closely into this graph, the difference between the revenue earned from just a 5% increase in retention rate is massive in 60 months. So, the insight you can glean from this graph with respect to churn is that if those 5% (between 5% and 10%) customers had got churned then the compounded effect of revenue loss over the term of 5 years would have been exceptionally huge.
This is the extent to which customer churn can impact your business. In the initial period (between 10 and 20 months) it is negligible but you can clearly see how the benefit of its prevention compounds over time (in the 5th year). And this brings us to another very important topic in the area of customer churn – customer lifetime value (LTV).
Customer Lifetime Value
Customer Lifetime Value is the total worth of your customer to your business over the entire tenure of their relationship. The cost of acquiring a new customer is always greater than keeping the existing ones. Hence, to drive growth you will always have to focus on increasing the LTV of your customers. The revenue generated during the initial period usually pays off to break even the cost of new customers’ acquisition. And the longer they stay, the more they start adding to the profit margin of the business.A common mistake new SaaS players make is that they give more focus on acquiring new customers than retaining the existing ones. This means they need to keep acquiring as many new customers as the number of customers they are losing to at least maintain the equilibrium.But since the cost of new acquisition is always higher than retaining the existing ones, they don’t realize that they are actually losing out in this game. They not only end up creating more pressure on sales but allocate more budgets on advertising, marketing, giving special offers and so on for attracting new customers.Hence, increasing the LTV by
should be one of the primary goals of SaaS companies. The simple formulae to calculate LTV is:LTV = (Customer Revenue per year) x (No. of years of customer relationship) – (Total cost of acquiring and serving the customer)So, the longer they stay, the more their LTV gets multiplied over time.
Impact on Valuation
Churn rate is one of the important parameters that decide how investors perceive your business.
consider customer churn to determine your company’s performance in the market and retention rate is essential in a SaaS company’s public valuations.Investors use LTV to predict how much profit your company would generate in the future. It is a clear indicator of the health of your business. High churn rate reduces the LTV and that can make your investors doubt the strength of your company and put on hold further investments flowing into your business.
Net Negative Churn
Net negative churn is like a dream come true for any SaaS company. It is achieved when the total additional
surpasses the revenue lost from cancellations and downgrades.
It is the situation where your current customers are spending so much additional money on your business that the revenue lost by churned customers is offset by it.
The main factors that help achieve net negative churn are through:
Customer’s renewal of subscriptionsCustomers switching from basic to advanced planCompanies upselling higher version products to existing customersCompanies cross-selling other relevant products to existing customers
So, if you have a net negative churn then it means you could still have an unfair advantage of growth in your business even without acquiring new customers.
Revenue churn is the direct measure of lost revenue as a result of customer churn and downgraded subscriptions. It can be measured in terms of lost monthly recurring revenue (MRR) or annual recurring revenue (ARR). It is often expressed as a whole number rather than in ratio.
The causes of revenue churn can be one of the following:
Lost contract or cancellationsDowngrades to a cheaper planCustomer acquisition by a competitorCustomer going bankrupt
Except for the last point, all the above three causes are in your control and largely depend on the quality of service you provide to your customers.
Customer Churn vs Revenue Churn
In order to evaluate the strength of your business you have to take into account both customer churn and revenue churn. Customer churn would tell you how good you are at retaining customers. Whereas Revenue churn would show how good you are at retaining your revenues.
They sound similar but are not the same. There are chances that your customer churn goes high for a period but the additional revenue spent by the existing customers would offset the revenue lost from the churned customers. In this case the revenue churn may remain the same or sometimes even negative if the additional revenue generated by the existing customers surpasses the lost revenue.
At the end of the day, your goal is to generate profits. So based on both customer churn and revenue churn, you can identify the area you need to prioritize to curb churn. For example, if there are 10 customers who are on a monthly subscription plan of $10 each and there is another customer on $200/month plan then your priority should be on serving that one customer who is generating the revenue worth more than the total revenue of other 10 customers.
Customer churn is more relevant in the beginning when every customer starts with an initial purchase of your product. But with time, each customer starts generating different revenues and hence the Revenue churn calculation becomes more relevant for each case.
Why do you need a Churn Predictor?
A SaaS is highly scaleable, meaning you have 100-2000 customers, compared to a service business where you are dealing with few customers and in constant touch with them, you are adding and dealing with customers with a much faster pace. This reduces the interaction and touchpoints that you have with each customer, from 100s of customers you want to know who is likely to leave so you can take action and waiting for them to churn is not a good approach to revenue of business.
How to build a SaaS churn predictor?
When churn happens the user has already made a decision to leave the software and it is already too late
A better way to save the user is predicting who is expected to churn and get them back before they leave
So if there is a way for SaaS companies to know which customers are on verge to leave they will be able to take action and get them back.
This is where a churn predictor comes in place.
What can be the characteristics of such a system?
Looking at the way how SaaS function, Off the shelf churn predictors will not work, or there cannot be one common set of parameters that impact churn for all SaaS companies.
The churn predictor works well when it is custom made for the SaaS, it has to be configured for a particular SaaS depending on the parameters.
So that means a churn predictor for a SaaS company A will be different for a SaaS company B, hence a churn predictor needs to be able to easy to configure for a company.
There are 2 steps to building a churn predictor.
Identifying the parameters that have an impact on churn by understand churn vs retain users
Developing a model that measures the predicts the current users who are showing tendency to churn
Every SaaS product has few metrics and these metrics can be divided into 3 zones
Red Zone - When the metrics are below some threshold and leading to a high probability of churn
Yellow Zone - when they are using it but have a probability of being an average user
Green Zone - when they are using it heavily and can be good referrals and power users of the application
We can call such metrics as Churn Impacting Metrics™
Every SaaS company will have a set of Churn Impacting Metrics™
The Churn Predictor should take these Churn Impacting Metrics and create an algorithm that predicts which user accounts are in the tendency to churn
This becomes input for Customer Support, Customer Success and Product teams to take action
The input to the SaaS Churn Predictor is CSV file with user account IDs and other Churn Impacting Metrics™
Output of the algorithm is the Retention Score™
Higher the score, higher the chances of retaining the user
Lower the score, means they are at risk of getting churned
This link can't be embedded.
Understanding the churn
The churn happens when the customer cancels the subscription, but it is already late when they cancel. There is a time between the signup and churn and this is the time which we have to keep the user
There is a behaviour pattern of churned and a retained user and for some set of parameters you will be able to see who is behaving like they will churn or they will retain with the app.
The first step towards increasing retention or controlling churn is to get these important metrics and we call it the churn impacting metrics
Step 1: Identifying the Churn Impacting Metrics™
To identify the churn impacting metrics, we make use of data to find those patterns, we have create 2 set of metrics
Global SaaS Specific metrics
Using the SaaS Churn Predictor we will be able to identify the set of metrics which have a higher impact on the churn score
There are 2 kinds of parameters that will help SaaS companies
Global parameters: Applicable for all SaaS
Personalised Parameters: SaaS Specific parameters
Number of Login by the user/account
Total Page Views
Number of Page views by the user/account
Total number of tickets raised by the user/account
Days to Last Login
Number of days to last login by the user/account
Number of open tickets for the user/account
Number of Bugs reported by the user/account
last Score provided by the user/account
Number of chats by the user/account
Days to last chat
Number of days to last chat by the user/account
Has the user completed the onboarding?
Expiry date of the current plan for the user/account
Which plan the user is on?
A unique identifier for the user/account
The amount charged to the customer monthly
The total amount charged to the customer
There are no rows in this table
Time to first value
Time spent on the app
There are no rows in this table
2. Data Preparation
In this section we cover the parameters that you can feed into the algorithm, following this section for more details
3. How to Run
In this section we talk about how we will run the system and also show the demo
4. How to Use it
In this section we talk about how one can use the churn predictor