A majority of companies across sectors like media, manufacturing, and financial services believe that telecommunications firms could be a better source for customer insights than even Google. Yet, almost half of the same brands are not ‘aware of operators’ ability to even offer these insights.’
This gap between opportunity and ability is one of the biggest challenges facing the telecommunications industry today. And it is not just a matter of perception. Over the next five years, telecom firms will have one of the largest CAGRs in big data and analytics investments. But these firms are getting much less out of their analytics investments than their digitally sophisticated counterparts.
Tellingly, telecom firms have yet to fully leverage analytics to halt or reverse their own decline in revenue, profitability and growth. An already shrinking core business has been under pressure from innovative over-the-top (OTT) services that have cornered a significant share of market and captured up to $386 billion in mobile voice revenue. At the same time, telecom firm service revenues and profit reflect little of the opportunity represented by skyrocketing data traffic and usage figures. The telecom business still languishes as a commoditized – to borrow a phrase from a global survey of telecom executives – “fat pipe of bits.”
Driving value through problem solving
So how can telecom firms quickly apply analytics to their own internal needs and start demonstrating bankable value to their customers? A holistic problem solving approach to analytics will enable the sector to extract sustainable value from its big data investments.
The holistic problem solving approach is tool, technique, and technology agnostic. Instead, its emphasis is on relevant problem solving built on a deep understanding of the strategic objectives and business context of each enterprise. This understanding informs the creation of customized solutions that are the right blend of strategy, design, analytics, process engineering and technology.
This context-driven rather than technology-driven approach can help telecom firms address some of their more pressing issues. Here, we’ll take a look at three examples:
- Customer churn
- New product development
- Network management
Customer churn is an effective proxy for a deficient customer experience. But it is only a symptom of a deeper problem that can be traced back to multiple sources, for example: the limitations of a telco’s traffic forecasting model or contact center script. Conventional churn models in the industry rely heavily on transactions and demographics, limiting the ability to effectively address symptoms, let alone identify and fix root causes. Current models also tend to be discrete, with functions like marketing, infrastructure, and product development dealing with churn in isolation.
A blend of analytics and process techniques can help operators trace churn back to the root cause and address it at source. A robust model using an array of structured and unstructured data sources enables more granular customer segmentation and more specific insights into customer expectations and service issues. It also enables more causal links to be established and addressed through process improvement techniques. For example, granular segmentation and insights can pinpoint exactly where an issue is occurring in customer support. A quick mapping of the process steps may then determine that the process is broken because there is no procedure for dealing with a specific customer request. From there, the process can be efficiently reengineered to fix the issue.
New product development
As churn becomes better managed, telecom firms can also start moving beyond customer retention to acquisition through products and services that maximize customer and enterprise value. Today, churn analysis may be used qualitatively to inform new product and service decisions. But firms seeking to truly differentiate themselves need to begin blending predictive analytics capabilities with new product and service development.
Predictive capabilities, or the ability to forecast and test scenarios of customer, competitor, and business changes to come, is becoming essential in an industry so prone to disruptive change. The ability to forecast business as usual trends and model the potential business impact of game changing innovations or consumer responses to price changes will enable firms to respond proactively to market dynamics.
Taking the approach further, firms can also begin analyzing prevailing product use patterns, defining core product attributes and test new combinations of product attributes that offers the best value. This analytical and structured product development method can be combined with micro-testing, or limited releases paired with rigorous performance analysis. Through this method, telecom firms can trial more relevant new products, assess financial performance, and quickly optimize product offerings before investing heavily in their development.
Turning to costs, telecom firms have major opportunities to reduce costs through more robust application of analytics, process, and management techniques in tandem. Infrastructure and network related operating and capital expenditures account for over 20 percent of operator revenues today. New forms of real-time analytics can substantially cut waste throughout the network but need to be blended with process simulation, risk management, and strategic asset management to realize this value on the bottom line.
A network requirements forecast can leverage similar techniques to those used for churn and new product development. What changes is the variety of data that reflects specific drivers such as customer data requirements or new technology profiles. Therefore, a forecasting model that can adapt to new information sources – such as a neural network model – is important in this arena. In addition, techniques from risk management and process simulation can enhance the forecast by looking at a variety of scenarios and assessing their likelihood of occurring in the future. This gives a complete picture of the network requirements over time rather than a static view.
Once forecasting is underway, the type of action that is taken is also critical. Optimizing the asset base requires a full assessment of when and how to phase in new technologies as well as when and how to phase out older ones. This raises a variety of needs for network modeling, process reengineering, asset management optimization, and total cost optimization. Drawing from a mix of analytical techniques alongside process, inventory management, maintenance and asset management among others can provide a comprehensive action plan to manage change while reducing overall cost.
These are just a few examples where the holistic approach to advanced analytics can be a transformative catalyst in the telecom industry. In these and other cases, simply introducing analytical models will not solve the business needs or yield the ROI required to justify telecom firms’ current and proposed investments in big data. Instead, telecoms need to take a broader view of problem solving originating with their strategic objectives and business needs. From there, choosing the right blend of management, scientific, analytical and design techniques will ensure actionable, monetizable, and sustainable business value is delivered.