Article Authored by Razorsight CEO Featured in Pipeline Magazine http://www.pipelinepub.com/Innovation
By: Charlie Thomas

The pressures facing communications service providers (CSPs) in the face of exponential data growth, from smartphones and increasingly connected devices to market saturation, declining average revenue per user (ARPU), and new competitors with significant economic advantages, have created a classic case of “Is the glass half full or half empty?”
The market opportunity has never looked better, and CSPs stand to profit handsomely from the mobile explosion and the monetization of “all things connected.” At the same time, the threats have never been greater, with innovative new technologies, including over-the-top (OTT) services and social apps that disrupt traditional telecom and cable models, putting CSPs at risk of being relegated to mere “dumb pipes” in the same way that Uber has upended the taxicab industry’s long-standing service model.
The winners in this new mobile world will be CSPs that quickly jettison old business models and rethink how they serve customers, both in terms of network architecture and a new economic model based on mass personalization. To execute this vision, they’ll need to leverage a major breakthrough in Big Data called predictive analytics, the brainchild of a new generation of data scientists. CSPs that follow this promising new path will redefine the industry and gain a significant competitive advantage in the same way that the Oakland Athletics redefined their 112-year-old “business,” baseball, using rigorous statistical analysis, as documented by journalist Michael Lewis in his 2003 book Moneyball(and the 2011 movie starring Brad Pitt).
Recognizing that the classic silo approach to products, markets and customers, not to mention the associated business intelligence (BI) models, is obsolete, CSPs are taking action, embracing a new approach to analytics that looks at those three main areas of interest through a 360-degree lens. These innovators represent a new wave in the telecommunications industry that will dominate the market from 2015 to 2020: by making predictive analytics their number one Big Data priority and capitalizing on new insights into customer wants, consumption and service challenges, they’ll be empowered to swiftly deliver premium services and mass customization on demand.
In the meantime the CSP “all-stars” are preparing for a whole new ball game on a much bigger field, one chalked by the next generation of smartphone and tablet users, who by 2020 will each generate and consume a zettabyte1 of information. Using predictive analytics, the market leaders in telecommunications will step up to the plate with the power to sift through data volumes at a level not previously witnessed.
As the first in the industry to apply the expertise of world-class data scientists to the challenges of the Zettabyte Age, Razorsight has been active since day one in helping CSPs find and leverage the long-sought ability to grow their businesses through the application of precise BI by customer, product and location. What we’ve learned and how we’ve applied this knowledge has set a new standard in the way that CSPs assess customer value, pricing, marketing, and retention as well as network investment. Carriers now have the power to measure any single customer’s direct contribution and to base business decisions on the current and projected value of said customer, the requirements for product development and the impact on network infrastructure costs.
The result: a true game changer that will positively transform the business of forward-looking CSPs. While their competitors sit on the bench, the industry’s “A” teams will consistently score home runs, an unprecedented batting average made possible by the ability to analyze off-the-chart data volumes and glean, with laser precision, the exact data required to monetize key, profit-laden insights into customers.
Big Data: the carrier’s perspective
Razorsight’s interest in predictive analytics for CSPs was prompted by daily interactions with industry leaders. Time and again we encountered a common thread of issues that troubled them:
- How much is a customer worth, and what drives each customer’s profitability?
- Which customers are likely to churn, and why?
- Which products is a customer likely to buy, and how will his or her decision impact the products’ lifetime value?
- How is my company’s network impacting customer satisfaction, and which customers should it address as priorities?
- Where will my company have capacity or configuration issues in the network, and how will they impact my top and bottom lines?
- How can my company best communicate with customers to drive response rates and lifetime value?
In theory, Big Data solutions should have addressed these issues all along, but the answers never emerged because old-school BI platforms weren’t designed to handle the complex needs of today’s Big Data. Unlike advanced analytics, which thrives on the massive amounts of data that are generated in the new “all things connected” mobile universe, BI can only partially supply the answers that CSPs need to know.
The irony is that some CSPs still aren’t aware of how they’re being shortchanged by their own Big Data platforms. Ask them if they have a Big Data analytics solution and they respond with an emphatic “Yes!” But when you take a close look at their platforms a different picture emerges, one with some obvious problems.
Because these CSPs still house their Big Data solutions in their IT departments, they’re forced to wait days, weeks or even months for answers to their questions, which are often based on hunches rather than facts, while siloed Big Data platforms are plagued by contradictory definitions of metrics that make it impossible to answer even the simplest question, e.g., “What is a customer?” Perhaps most frustrating of all, the personnel in charge of operating the siloed Big Data platforms are forced to manually link sources within the Big Data platform to gain a consolidated view of the customer, leading to long delays, abundant errors and multiple versions of the truth.
The clock is ticking. Under intense pressure from growing volumes of data on one side and profit erosion driven by plateauing revenue and increasing costs on the other, many CSPs continue to lose ground because the answers to their most fundamental business questions remain a mystery.
Where predictive analytics delivers results
- Enterprise-wide value. Predictive analytics organizes complex operations and customer data into a multidimensional model to facilitate sophisticated analysis so that users in any of a CSP’s departments can directly and easily consume data.
- Deep analysis. It also incorporates intuitive, three-dimensional data visualization, granting CSPs the power to drill down for rapid-fire analysis and problem identification.
- Data integration. By pulling from hundreds of sources, including billing, accounting, network feeds, inventory, customer relationship management (CRM), customer care and activation, enterprise resource planning (ERP), and pricing, predictive analytics reveals what a customer is worth at any given moment in time, whether now or in the future, thus completing the current “mission: impossible” of Big Data.
With these insights CSPs can make informed decisions that precisely determine the lifetime value of a customer, the likelihood of churn, when (or whether) to invest in product development and loyalty programs, and how much to invest by segment and individual customer, as well as the impact on the network, the expected return on investment (ROI) and profit potential.
Finding the perfect predictive-analytics opportunity
The challenge of where to start can slow a CSP down even when it recognizes the need to take action on its Big Data issues. Unfortunately, this inertia is exacerbated by vendors that flood the marketplace with inflated promises of analytics capabilities that generally fall short of what a true predictive-analytics application can and should do. To guide your selection process and avoid potential pitfalls, you need to know what to look for right off the bat:
- Flexibility and scalability. Insist on a cloud-based approach, as it provides maximum flexibility, meaning you can start small and expand, or contract as your needs change, without the risk of upfront capital requirements or licensing costs.
- Industry experts. It is essential that your predictive-analytics application be backed by business experts with deep experience in the telecommunications industry and data experts who understand not only how to properly normalize data from the full array of data sources but also how to optimally leverage predictive-analytics engines.
- The right questions. The application’s system and design must allow you to ask the right questions and perform specific analyses for your business via multivariate statistical capabilities.
- Operationalization. It must also be capable of becoming an integral part of your everyday business, and it’s imperative that such an application be fully operationalized to capture the full value needed by all teams across your organization.
The top analytics priorities for your chosen application should include:
- Statistical analysis. The application must use mathematical algorithms to identify hidden relationships between events, people or actions. This can take the form of root-cause identification, social-network analysis, behavioral segmentation, or semantic analyses.
- Propensity analysis. Statistical algorithms are essential for showing what will happen to an entity in the short term (e.g., within 90 days of today). In addition, propensity analysis can aid in the understanding of the behavioral drivers of such predictions.
- Forecasting. This capability uses econometric and statistical algorithms to forecast what will happen next week, next month or next year. It also enables “what if?” analyses to help with scenario planning.
- Optimization. This one leverages advanced statistical algorithms to identify the best possible outcomes when given a specific set of constraints.
The ideal predictive-analytics application will take your company to its endgame through crucial insights and data analyses that enable value-based decisions honed to customers with the greatest current, and forward-looking, profit potential. The solution you select should be able to define organizational goals and monitor them via key performance indicators (KPIs) and scorecards, visualize and aggregate metrics into an overall score and make it possible for anyuser in any department to review the data by using the application’s dashboards.
Ease of access across the organization is a must. Predictive analytics empowers managers and line organization users, whatever their specialty or interest, with near-real-time actionable intelligence that provides a laser focus on a company’s most valuable customers—by name, by the products they use and where they use them. Moreover, the application goes one step further by applying capabilities that predict what these customers will want to buy “down the road,” what that will cost in terms of personalized product development to meet their needs and what the impact will be on the network.
With this pinpoint-accurate data at hand, everyone sees the full picture and is in a better position to contribute to value-based decisions on which actions to pursue—at the appropriate level of investment, of course—in order to fuel savings and accelerate profits.
The ideal predictive-analytics application will take your company to its endgame through crucial insights and data analyses that enable value-based decisions honed to customers with the greatest current, and forward-looking, profit potential. The solution you select should be able to define organizational goals and monitor them via key performance indicators (KPIs) and scorecards, visualize and aggregate metrics into an overall score and make it possible for any user in any department to review the data by using the application’s dashboards.
Ease of access across the organization is a must. Predictive analytics empowers managers and line organization users, whatever their specialty or interest, with near-real-time actionable intelligence that provides a laser focus on a company’s most valuable customers—by name, by the products they use and where they use them. Moreover, the application goes one step further by applying capabilities that predict what these customers will want to buy “down the road,” what that will cost in terms of personalized product development to meet their needs and what the impact will be on the network.
With this pinpoint-accurate data at hand, everyone sees the full picture and is in a better position to contribute to value-based decisions on which actions to pursue—at the appropriate level of investment, of course—in order to fuel savings and accelerate profits.
Predictive analytics at work
Early movers have been quick to embrace predictive analytics, including a large tier 1 provider with a solution now in production that’s already realized significant gains in operations and its bottom line.
Before adopting predictive analytics, this CSP had faced issues common to all in the industry: it needed to leverage Big Data to drive profits but was swamped by growing volumes of data, plus it was held back by stovepiped systems and multiple tactical-analytics capabilities limited to a single perspective. Data gathering proved to be time-consuming and resource intensive, and the company faced numerous internal obstacles in its analyses of the results, including high-level allocations that didn’t represent actual activities or metrics.
The output of old-school analytics was too fractured to indicate a clear path of action. So, after years of relying on a platform that failed to provide the most essential data—profit value by customer, product and location—the CSP opted for a new approach to meet its objectives: Razorsight Predictive Analytics.
Razorsight’s team of data scientists deployed predictive analytics as an overlay hosted on a secure, private cloud, and today its application is helping the CSP rack up achievements it once thought were impossible, successfully integrating over 100 data feeds from a variety of source systems (CRM, ERP, etc.) to present a single view of each targeted customer by product and exact location. The overlay deployment allows data transmission and ingestion in a loosely coupled model, eliminating the risks typically associated with tightly coupled integration models. All workflow is hands-free automated: data reception, tracking, ingestion, and processing are centralized in the Razorsight Predictive Analytics platform to ensure faster profiling, analysis and deployment. The cloud-based model of the application has also completely wiped out previous costs related to hardware, software and IT.
The CSP can now immediately access an exact view of its highest-value customers based on near-real-time data, leading to millions of dollars in savings and improved EBITDA (earnings before interest, taxes, depreciation, and amortization) profitability. With predictive analytics at its disposal, the company is setting new records in efficiency that positively impact legacy-product sunset, targeted promotions, pricing refactoring for certain product lines by market, and network consolidation. The payoff is measured in:
- improved management across periods of product growth, harvest and sunset;
- faster response times to customer needs that might be unleashed by an advance in technology or a demand for new services;
- an established lead, fueled by actionable insights into customer profitability and price sensitivity, in an intensively competitive market;
- an improved financial performance triggered by new profits from the CSP’s highest-value customers, products and markets;
- the added value of an aggregate view across its organization.
Is predictive analytics the future?
Looking ahead, many in the telecom industry believe that predictive analytics focused on customer, product and location data will become the new norm, a capability that all CSPs will demand. And to understand why it will put smart carriers in the Zettabyte Age’s big leagues, one need look no further than the customer.
In a market saturated with look-alike competitors and me-too products, customers seek true differentiation between their service providers. Increasingly, “meaningful separation” from competing products rests on a CSP’s ability to create personalized offers that deliver exactly what each customer demands. Predictive analytics arms that company with the precise data it needs to craft tailored products that match and serve its customers’ needs. Just as important, this “harbinger” application informs the company of the best ways to market products, ensuring rapid acceptance and a loyalty-boosting, positive experience for customers at the right investment threshold, which is critical for establishing profit optimization.
Predictive analytics achieve what Big Data was meant to do all along, quickly simplifying massive volumes of data into usable bytes that put CSPs in sync with the individual who matters most when it comes to financial performance—the customer.
Full Article: http://www.pipelinepub.com/Innovation/big_data_analytics

________________________________________________________________

2013 Berkshire Annual Meeting – CenturyLink Center

Ajit Jain and Warren Buffett
Charlie Munger and Warren Buffett