It’s no secret that competition among communications service providers (CSPs) has intensified. And customers have learned to churn, that is, blithely switch from one provider to another. That’s revenue walking off the balance sheet and big data analytics can stop it, writes Syed Mahmood, senior product marketing manager – analytics for Tibco Spotfire
Over the years, CSPs have responded to churn through strategies both trivial and> thoughtful. Marketing teams scramble to keep up with evolving customer segments. The urgency is even greater in call centres, where customer service representatives want to know what offers to present to a customer while they’re still on the phone. Over in network operations, managers try to decide where to invest in new capacity for the future.
Instinctively, organisations search for clues in their data, of which there’s no shortage. Analysts scour customer behaviour and the concerns and needs expressed to customer service representatives. Call records, network data, social media as well as many other sources, internal and external also give clues. But even with all relevant data, managers soon find their analysis tools’ limitations. They simply can’t answer critical questions. Old assumptions have resulted in a rigid data structure and reconfigurations often require IT assistance.
Churn carries some serious threats but also opportunity:
Multiple service revenues are lost when today’s bundled customer cancels. The frustration a customer may have with their phone service could mean lost revenue from canceled television, internet and other services.
Social media sets off a chain reaction that magnifies customer experiences. The one-to-one conversation between friends and acquaintances has ballooned into the hundreds or thousands on Twitter, Facebook and elsewhere.
Churning customers provide insight, if it can be captured. At the moment a customer closes an account, the CSP records reason codes. Though not always reliable, reasons to switch give clues to network weakness and vulnerability to the competition.
The soft stuff really counts
Trivial offers can be infuriating to anyone who has suffered a series of disappointments, and now only craves serious attention. Customers rate “customer service, honesty, and trust” just below strong network performance when it comes to brand allegiance, according to a 2012 study “Customer Loyalty” by PricewaterhouseCoopers. The trouble is that perceptions vary. What one segment perceives as attentive service, for example, may seem annoying or intrusive to another.
Much can vary by a customer’s age group. Older users tend to value consistent customer support even more than younger customers do and are less impressed with incentives. They are also much less likely to switch than younger customers, viewing it as more of an inconvenience than an opportunity.
Fighting churn one caller at a time
Some of the most important battles against churn occur in call centers. There, each customer service representative makes immediate use of any knowledge – any that’s available at a glance.
Those few moments are critical. Customers who call with a problem and hang up with a satisfactory solution are more loyal than those who’ve never had a problem at all. Satisfy the caller’s need and the caller will keep coming back – but only if the representative has a full view of the customer and knows what can be offered.
Angel or demon? A delicate issue in this business is that some customers just aren’t worth saving. The opposite of the highly profitable angels are the demons. The demons are not only less profitable than average, the provider actually loses money on them.
A 2011 Pitney-Bowes whitepaper, “Customer Centricity in the Telecommunications Industry,” proposes estimating each segment’s lifetime value. The formula involves ARPU (average revenue per user), churn and retention rates, CPGA (cost per gross add), revenue and other common data. Both new and traditional metrics are ultimately just trivia unless they’re actionable and such action usually takes place first in the call centre.
Recommendations
- Make self-service data discovery a part of everyone’s job. Make it easy to use any source of data necessary without delay. Give them a tool for dragging and dropping sources into a data mashup, not a form for asking and waiting for IT.
- Provide visualised data. Help ordinary business users to intuitively grasp statistical significance without requiring deep knowledge of statistics.
- Utilise the power of predictive analytics. Advanced predictive models help to pinpoint opportunities and risks by uncovering hidden relationships, patterns and emerging trends that might be otherwise go undiscovered.
- Encourage the proliferation of insights. Most organisations contain enormous knowledge and insight within them, but much of it is stuck in silos. Contextual collaboration on an analytics platform extracts it and puts it to use throughout the organisation.
Real competitors don’t wait for that fatal moment of customer defection. They look far into the data to spot triggers and events that warn of imminent churn, realizing real insight into the unknown. Analysts integrate the data they need, as they need it without help from IT.
The modern analytic platform propagates insights in role-based and secure views wherever they’re needed, and wherever analysis might incite collaboration. The quality of interaction rises and loyalty grows when the company shows it truly knows what each customer needs.
