Building a better business one customer at a time

Customer churn is one of the most pressing issues the telecommunications industry faces—and it affects all types of carriers from cable operators to mobile service providers. According to a study by Bain & Co. Inc., companies can boost revenues by as much as 85% if they can retain only 5% more of their best customers. 

In the economic downturn, telecommunications service providers continue to focus on reducing churn while placing more emphasis on increasing the profitability of existing customers. The tools to make existing customers into more profitable customers lies in the data a service provider already has in its own environment. The key is to collect the right data to determine the “why” behind customer churn or behind customer buying behavior.

When it comes to customer data, collecting everything means nothing. According to industry research firm Gartner Group, analytics have increased in importance as enterprises recognize their potential for alleviating the paralyzing condition known as infoglut—an overwhelming information and data overload. Infoglut results from a mismatch between an enterprise's ability to acquire and collect information and its ability to read, understand and digest the data it has gathered. Enterprises may pay for their failure to invest in analytics with decreased productivity and inferior decision making.

The key to putting predictive analytics tools to work successfully is on a one-to-one basis, targeting the individual customer, not the customer base as a whole.

Predictive analytics—using models to predict future events from today’s data—is the killer application to help service providers not only survive infoglut, but also realize value from the sea of data gathered, especially as providers roll out advanced networks and offer more services to entice and retain customers.

A new generation of tools that deploy complex algorithms and models will enable operators to predict which customers are at greatest risk of churning and how to increase subscriber value through individualized, optimized product and service recommendations.

The key to putting predictive analytics tools to work successfully is on a one-to-one basis, targeting the individual customer, not the customer base as a whole. Often, those service providers who already analyze customer data stop at general customer base trend and segmentations analysis.

Analytic data models and human analysis combine to return recommendations on upselling and retaining individual customers drives real value for providers. To accomplish these goals, one of the first distinctions to understand is the difference between customer value and customer profitability.

Customer Value Vs. Customer Profitability

Customer profitability and customer value are two distinctly different issues to measure.

Subscriber or customer value takes into account the entire relationship, including factors such as customer referrals. Profitability is purely financial—how much revenue is the customer bringing in versus the costs of delivering the service?

To gain a more accurate view of customer profitability, the comparison of revenues against costs takes place not at the service level, but down to the individual subscriber level. The answers about customer profitability exist within the information that providers already collect within front and back office systems, including billing engines, data warehouses, data mediation applications and so on.

Some typical costs that impact customer profitability include call center support, churn and package churn and network capacity used to provide the service.

For example, a text message may have a 90% profit margin for the carrier, versus voice, which may yield a 50% to 60% margin for the carrier. To increase customer profitability, service providers can determine the value of the message to the customer to market services to those users who are willing to pay a higher price to send a text message, resulting in more profit for the provider and shorter usage of the network.

Service providers can also steer customers toward more profitable ways to pay. A prepaid customer for example, who produces $15 per month in revenues actually could be more profitable than a post-paid customer who pays $40 per month for services that cost more to deliver.

Providers can also predict which new service plans may be more profitable in the long term. As customer service representatives (CSRs) make offers to customers to upgrade to rate plan A, predictive analytics help the provider understand that plan A has a 15% chance for customer acceptance and offers less revenue up front to the service provider. Rate plan B offers a 30% chance and carries a higher price, which brings more revenue to the service provider, faster. However, if the customers who sign on to plan A actually stay with the plan longer than those who sign up for plan B, plan A might be more profitable over time although the immediate profit from it is smaller.

How can providers draw these kinds of conclusions about the products and services that they offer in relation to a customer’s potential buying behavior? Predictive analytics provides a solution when supported with the right data.

Making Better Business Decisions

The most reliable data for predictive modeling is available from operational, automated systems. For telecommunication carriers, the billing system is vital, followed by network management data and call center tracking information.

The most reliable data for predictive modeling is available from operational, automated systems. For telecommunication carriers, the billing system is vital, followed by network management data and call center tracking information. Data that must be correct for the proper operation of the business is more effective than subjective or intermittent data.

How does a provider determine which existing data will help address specific business problems? Identify the correct data to get to the “why” of the problem, then determine if the data that will provide the answer is already being captured. For example, if a provider wants to know why customers in a certain zip code are churning at a higher rate than other customers, is location-specific data such as record of service outages available?

Using data modeling to define what data is relevant and important can help service providers resolve questions such as these. The combination of automated modeling and human analysis can categorize the data gathered to give a more detailed focus into the critical indicators of “why” a customer does what they do.

Common questions that service providers should be asking:

• Am I accurately capturing, on a per-subscriber basis, all relevant revenue and cost events?

• Am I capturing information necessary to identify different types of churn; e.g. structural, voluntary, and non-payment churn?

• What can the data tell me that I can act upon?

• Can I tell if my business is changing based on my data?

• Am I using the right data?

• When looking at how to keep my customers—am I recording (properly) why they left?

Taking Action on the Data

To take action on existing data, service providers must be able to operationalize the data, or make the information actionable by CSRs and other customer touch-points during live customer interactions. Even the most insightful data reports can’t save an individual customer from churning—if the CSR can’t access the data in an easy-to-use format and put the information to immediate use. Until the operationalization of the data takes place, a service provider has yet to take the last and most important step of any customer analytics program. 

Special care must be taken to account for the business context of a predictive analytics application. For example, a churn application should treat high-value customers differently than customers who are non-payment risks, even if both have similar churn scores.

False Predictors

False predictors are unique to predictive modeling, and are a by-product of the nature of modern data warehouses. Predictive models rely on past patterns to predict future patterns, yet most data sources provide only a snapshot in time. A false predictor arises when fields in the source databases falsely appear to predict the model target, when that relationship is actually an artifact of the data source.

For example, most telecommunications databases will have a list of equipment items in the subscriber’s possession such as mobile phones and cable boxes. When a subscriber churns, most companies delete or reassign all equipment records belonging to the customer. This makes count of subscriber equipment records a false predictor, because if the subscriber has no equipment, they have already churned.

The false predictor actually throws the model off; predictions made on customers with equipment records greater than zero will be worse than they should be. However, simply discarding this field from the model is less than ideal. For instance, the count of equipment records before a subscriber churns is almost always useful for predicting churn. Typically, the more equipment a customer has from the provider, the less likely they are to churn. In practice, false predictors are very common, and eliminating them can require a combination of both human and software intelligence.

The Future of Analytics

As the global marketplace becomes increasingly competitive and retaining the best, most profitable customer relationships a service provider has becomes more important than ever before, predictive customer analytics will see more widespread use.

Since the variety and depth of data from these exercises can be extremely complex, good analytics products won’t sacrifice accuracy and the complexity, but they will make it easier for the end users to understand the results. The output of analytics recommendations must be designed for the CSRs taking live action on analytic data—and must apply to any customer contact level within an organization. In addition, providers will turn to analytics to measure a potential new customer’s credit risk based on their credit history before allowing the customer to sign up for services.

Providers will also move away from analytics systems that provide data without any context or wisdom.  Instead, new analytic systems will explain each prediction in the context of an individual customer so that marketers, call center operations, sales and others can expect to understand why a model has scored a customer's credit, churn, fraud and more in a certain way.

Finally, new tools will give marketers the ability to input fuzzy knowledge about market forces such as competitive threats, revenue and loyalty directly to operational analytics engines—and the results will immediately affect customer retention and profitability.

Analytics tools can move not only into high-level marketing strategy, but also carry that strategy into day-to-day, bread and butter tactics. Providers then will know what's working and what's falling flat.

---

Richard Wolniewicz is Vice President-Engineering, CSG Analytics for CSG Systems, Englewood, CO, a vendor of customer care and billing software and services for the telecommunications industry. He can be reached at richard_wolniewicz@csgsystems.com.

Visit CSG Systems online.