What makes your customer stay or leave?
Do you want to know what factors keep your customers satisfied, or why your customer is dissatisfied? In this blog, I’ll open the core of customer churn analytics, and the benefits it offers to companies. It’s a question of the ability to predict and analyse things that weaken the customer relationship, or vice versa, which factors behind increased satisfaction come up in long-lasting customer relationships.
It’s important to identify dissatisfied customers before it’s too late. Satisfied customers are the best advocates of a company, whereas those who leave take away not just their share of the turnover, but also a message of their dissatisfaction with the service they have experienced. Do you know how loyal your customers really are, and how that loyalty could be strengthened?
Customer churn analytics is most straightforward in companies with contract-based customers. These are, for example, companies offering maintenance services, electricity companies, insurance companies, banks, gym chains, telecommunications operators and the providers of entertainment services, to mention a few. Customer churn analytics also works without a contractual setting if a customer can be identified by some other means, for example through registrations in online stores. Churn analytics can be applied to both B2B and B2C markets.
When analysing customer churn, at the same time we also build up a view of the differences and needs between the customers. If different customer segments are found, offering a targeted services or new products might be worthwhile. Let’s take an example of car servicing. Does servicing always according to the same checklist, spare parts and price list drive away the owners of an old car from recognised make garages, because they are afraid of high costs? Could motorists be offered servicing plans customised according to their usage habits, service history and make-specific characteristics? Would owners be interested in the additional services related to the safety of the car, insurance policies, or trading in cars? A service that hits the spot could increase the customer satisfaction and loyalty to a wholly new level.
A lost customer is always a cost
Sometimes the loss of just one customer can create a significant dent in the turnover. When a large company terminates its staff restaurant service contract, we’re talking about a loss of a different scale than when a customer terminates the family’s phone and Internet subscriptions. But if the number of these subscription terminations begins to pile up, it will show directly and negatively in the turnover.
Acquiring new customers is not free. The money invested in customer acquisition has its payback time, and building trust does not happen overnight. Customer churn accumulates in the longer term into considerable sums of money, so it’s necessary to ask: Could something have been done to keep the customers? What does acquiring a new customer cost and how laborious is it?
I’ll use the following example from maintenance service business. Companies providing maintenance services typically have customers with contracts lasting many years. In this business the most obvious form of analytics is utilizing sensor data for predictive service and maintenance analytics, which ensures that both the service engineer and the spare parts are in the right place at the right time. Servicing has to be timed and focused correctly to ensure safe and reliable use of the equipment. However, in addition to equipment logs and sensor data, often some data related to the customer itself is also available, which can be refined to find indicators of the customer satisfaction. When it is analysed what kind of common factors there are behind both terminated and renewed contracts, valuable insights are gained for developing service models, improving customer experience and identifying the ’pain points’ in the customer relationship.
How often has a customer needed technical help, and how well was the support taken care of in the end? Have the ordered spare parts been delivered according to the agreed schedule? How often has the customer had outages with their equipment? Payment behaviour, length of the customer relationship, customer survey results, or frequency of contacts with the customer can also serve as signals about customer loyalty, or the increasing desire of the customer to terminate the contract.
I recommend that you think about your business through these questions: Is customer churn causing concern? What should be done better?
Get going gently and try things out
We offer companies an easy and affordable way to start trying out customer churn analytics. A low-threshold approach is the best way to proceed, and detailed project plans are not required. We can look together at what kind of data can be found for your case, and what we could get out of it.
Successful analytics is a team game. In addition to someone skilled in analytics, the company’s own business understanding is also essential in concretizing the objectives, and finally with the aid of a data warehousing expert, we can design a data set relevant to the problem. After quality inspections, data is refined by constructing features from it, which are assumed to have forecasting power in differentiating loyal and departed customers. The differentiating ability is tested through machine-learning methods, and when evaluating results in the end, we draw conclusions based on both analytics outcomes as well as such tacit business knowledge that cannot be captured by data.
If the results of the pilot stage described above show promises, analytics can be productised as a part of daily decision-making. Then the data flow required by the model is automated, and the model is kept up to date by ’training’ it at regular intervals. We can design an easily comprehensible dashboard from the results, and train the end-users. Then everyone has easy access to the same and up-to-date data-based view for nurturing and developing customer satisfaction.
A positive cycle brings growth
There is nothing mystical in analytics, and it doesn’t require magic tricks. Need for analytics originates always from the business needs. Trying it out does not require commitments or investments in any specific technologies; you can get going in an agile way with open source tools and limited amounts of data. Instead, analytics requires an open-minded attitude, a little effort in digging out representative data, and the management support in getting into uncharted waters.
Begin with a light-weight trial and assess the possibilities. I recommend choosing a partner who has the ability to grow along with your needs. Remember that a successful analytics try-out can start a positive cycle which expands and creates more growth opportunities for your company!
Download our solution paper and read more on customer churn analytics and the benefits it brings to your business.