Retaining existing customers is almost always cheaper than acquiring new customers and it has never been more critical to understand where there is the risk of churn in your customer base so you can take targeted action to prevent it.
Following a recent ECI Unlocked webinar from data and analytics consultancy, QuantSpark, we have pulled together our practical top tips for how to use analytics to predict and prevent churn:
1. Before you start any analysis, make sure you have access to the right data
One of the first actions companies need to take is to check the customer revenue data is accessible, accurate and in a useable format. This could mean reviewing invoice-level data to understand data quality, which can take time, but the accuracy of the output will only be as good as the accuracy of the input data.
Building a view of recurring revenue by customer and service, and enriching it with information such as customer size, sector, tenure and satisfaction metrics will be important for segmenting your customers and as input variables for predictive analysis later on.
2. Next you need to have a clear and consistent definition for churn in your business
It is not unusual for Finance to be working with a different definition of churn to Customer Success, for example, but having one common understanding of what constitutes a churn event in your business is key.
Some example questions to consider:
- For a customer who pays annually in advance, at what point in time do you consider them to have churned? The point they notify you they will not be renewing their contract or the point where the subsequent invoice is not paid?
- Are you focused solely on customer-level churn, or should you also include product churn, where a customer drops one or more products, but retains another?
- How do you treat customers who leave or pause for a period of time before returning?
3. Decide on the level of churn analysis that is appropriate for your business
a. Churn Visibility
What is churn visibility analysis? Simply providing visibility of the level of churn in the business and tracking this over time.
Including this information in Board packs and making it accessible for people across the business using self-serve dashboards will help push churn up the agenda and ensure it gets the right level of focus. Often shining a spotlight on churn and galvanizing the team around the need to manage it is enough to have a positive impact. It will also allow you to see the impact of business decisions that you make on customer retention.
You may be thinking, what are the best churn KPIs to include? Here are some standard definitions that you may want to consider:
- Revenue churn: Percentage of opening revenue lost from cancelled contracts
- Gross revenue retention: Percentage of opening revenue retained after removing cancelled contracts and downsell
- Net revenue retention: Percentage of opening revenue retained after removing cancelled contracts and down-sell, and adding upsell in continuing customers
b. Customer Segmentation
What is customer segmentation analysis? Analysis to understand which customer segments have historically had the highest levels of churn. This can be based on end sector, products/services taken, tenure, customer satisfaction metrics or even how they interact with your business. Start by building a set of hypotheses for the drivers of churn through qualitative discussions with accountant managers or other stakeholders to be tested with this analysis in order to focus your effort in the right areas.
This analysis can help to identify higher risk groups that should be monitored more closely, or process improvements that could be made to better retain certain types of customers. A positive is that this segmentation can be done without complex code in programs like Excel.
c. Predictive analytics
What is predictive analytics? Using advanced analytics to identify the relative importance of different customer characteristics and triggers on the likelihood of a customer churning, and using this to give each customer a risk score based on their predicted likelihood of leaving at a given point in time.
This analysis can then be combined with lifetime value analysis to understand the relative value of different customers to allow you to carry out targeted account management actions at a customer level, to prevent the highest value, highest risk customers from churning.
There are also a number of off-the-shelf tools for monitoring customer health and other leading indicators for churn and using this to drive alerts and automated actions to engage with at-risk customers. One example is Gainsight, which is currently used by two software businesses in our portfolio, Mobysoft, a provider of predictive analytics into the social housing sector, and Ciphr, the cloud-based HCM software.
4. Whatever analysis you do, bring the business on the journey with you
To bring people along with you, you need to be able to explain:
- Why is managing churn important for your business?
- What would be the impact of reducing it?
- How will it help you hit your business goals?
Involving stakeholders from across the business early so they understand the answers to these questions and feel ownership of the work will be critical to its success. Your team are talking to customers regularly – use their insight to guide your analysis.
It’s important to keep your analysis simple enough to explain to anyone in the business. It can often be better to sacrifice a small amount of accuracy to use a model that is simple enough for stakeholders to understand and buy into. It can be easy to overengineer the solution by including too many data points or trying to answer too many questions. Start small, demonstrate success early and then build from there.
5. Act on the findings
Understanding where to focus is only one piece of the puzzle, the real value comes from the actions you take as a result.
A good example of this was with Moneypenny, a global third-party answering service. Our Commercial Team worked with management to assess which customer segments were driving the highest churn. This highlighted that smaller customers in the first 6 months were particularly at risk and led to a number of operational changes in the business. The team changed the way they incentivised the sales team to focus on year one retention, changed the way they managed the year one relationship with the client to improve engagement, and changed the focus of their marketing spend. This helped drive a 30% reduction in revenue churn between 2020 and 2022. They have since gone on to build a churn predictor dashboard to flag at-risk customers.
6. You can leverage the insights from the churn analysis to support other parts of the business too
Once you have a clear view of what is causing customers to churn you can develop new products and services to solve for these needs, or make process improvements, for example around how customers are onboarded and engaged with over time.