Only 5% of customers with a problem will contact your organisation. The rest will just leave. Customer churn prediction based purely on analysing data from the contact centre or from customer satisfaction and NPS surveys, will provide a very limited view of what’s going on. Your analysis and churn prevention strategies will be skewed to a very small segment of your customers.
The impact of customer churn is having a devastating impact on the bottom line for many organisations. Taking the Australian health Insurance sector as an example, the industry experiences around 10 percent of policy holders switching providers each year. According to health insurance industry consultant, Avnesh Ratnanesan, churn in that industry alone is putting $2 billion of revenue at risk on an annual basis.
Though brands recognise the financial benefits of reducing churn, the strategies and tools they deploy to prevent it are largely ineffective. The main problem is due to the data they use to predict churn is drawn from a limited range of sources and customer interactions.
Recent research published in Forbes highlights, “Companies mainly rely on post-call surveys to understand their customers. Historically, the response rate for customer surveys has been between 10-15% for most companies. With such a low response rate, it is impossible to draw accurate conclusions with respect to the whole customer base”.
The squeaky wheel gets the grease
There’s a tendency for brands to be reactive when it comes to customer service. They only respond to a customer’s issue when the customer complains to the contact centre. If only 5% of customers are making the effort to complain and have their problems resolved, that means 95% of customers are not receiving any service or help to resolve their problems.
More importantly it makes it almost impossible to measure and understand the full range of issues that are driving customers to churn. If you don’t understand why they are leaving how can you possibly devise strategies to prevent it.
Many of the existing methods and tools are simply unable to discover the root causes of dissatisfaction and churn. Data from surveys can be plagued by biased response samples, low coverage and variations in customer interpretation of the scales typically used in surveys. Speech analytic tools used in the contact centre to analyse customer’s tone and language are limited by the accuracy of speech mining and its ability to deliver true insights. These methods can reveal overall trends in the customer experience, but they can’t explain the ‘why’.
Seeing the bigger picture
Recent advances in AI and big data analytics has given companies the ability to analyse all data concerning interactions with their customers. Data related to operations, billing, onboarding, the sales process, product usage, account adjustments, discounts applied and of course the call centre, NPS and customer satisfaction surveys can all be analysed.
AI enabled analytics can sift through and analyse much larger and more complex amounts of data more quickly, compared to traditional methods. It also provides significantly more insights and opportunities, including ones no one even realised existed.
AI offers the ability to track the health and performance of each relationship a brand has with a customer. This insight can be used to identify the customers that are most likely to churn and identify the issues responsible for the dissatisfaction. Rather than analysing the feedback from a small percentage of your customers, you can assess and analyse your entire customer base.
