Delivering Big Data Governance in Telco: Interview with Sunil Soares of Information Asset, LLC

Big Data Governance is critical to Telecoms providers Big Data Governance is critical to Telecoms providers


“inconsistency in product naming and hierarchies causes a number of issues for telcos that have multiple silos or that have grown through mergers and acquisitions”

— Sunil Soares


The Telecoms industry is undergoing major upheavals across the globe as the demand for information services and technology marches forward. This demand creates vast quantities of data that demands comprehensive data governance controls but how should these be delivered? What data falls into the “Big Data” category? What are the drivers for governance of this Big Data?

Sunil Soares, founder of Information Asset, LLC and one of the world’s leading expert practitioners/ authors on the subject of Big Data and Data Governance recently took time out to answer the Big Data questions that are challenging Telecoms leaders today.

Dylan Jones: In terms of “Big Data”, what type of data within a modern telecoms company do you feel falls into that category?

Sunil Soares: Telecoms companies have assembled some of the largest collections of Big Data. Based on work that has been done by Arvind Sathi at IBM and myself, this Big Data can be broadly classified into the following types:

  1. Network events — Telcos have been collecting enormous volumes of data from network equipment that provides granular insight into how the network is being used. Once correlated with users, network data such as dropped calls provides valuable information about service quality and usage.
  2. Call/usage detail records — Telcos have traditionally collected a fair amount of information about calls such as origination, termination, and duration for voice circuits. As data and video products have evolved, the traditional CDRs have been extended to xDRs to also include information about the use of data and video. (The x is a variable that can be used to represent voice, data, or video information.)
  3. Location — Telcos have detailed location information on their subscribers. For wireline subscribers, this is typically the location of the digital subscriber loop endpoint. For wireless subscribers, this would be the location of the wireless device, typically computed via triangulation of cell tower locations serving the subscriber. With near-field and GPS-capable phones, this information can be accurate to within a couple of meters.
  4. Web traffic — Devices store a lot of information about web traffic, most commonly in web cookies left on the device.
  5. Channel clicks — Cable operators’ set-top boxes provide detailed information about subscribers’ channel selections. As other Communications Service Providers offer cable services on their fiber networks or Internet protocol television (IPTV) services, they will also collect information about content viewership.
  6. Social media — Telcos have been monitoring comments posted by subscribers on social networking sites such as Twitter and YouTube in response to new product offers or, in general, about customer service. Telcos are also beginning to use social networking sites to interact with their subscribers as they build their brand images and offer new products.

Dylan Jones: What are some of the most critical corporate drivers you see in telcos, what is pressing heavily on the minds of executives right now?

Sunil Soares: There are a number of critical business drivers for Data Governance in telcos:

  1. Improve the overall customer experience by establishing a single view of the customer across multiple silos
  2. Support product standardization initiatives across the enterprise
  3. Enhance the effectiveness of performance management, capacity planning, and location-based services by increasing the reliability of network inventory and topology data
  4. Secure access to sensitive data such as xDRs and location information
  5. Ensure consistency of critical business definitions

Dylan Jones: Given these pressures, how can data governance help? Can you give an example of where you’ve seen improvements drive benefits to the executive agenda?

Sunil Soares: Let us consider an example of how data governance can support product standardisation initiatives at a telco.

In a typical telco, one division reports revenues separately by red and blue mobile phones of a certain make and model. Another division does not break out the color separately. The inconsistency in product naming and hierarchies causes a number of issues for telcos that have multiple silos or that have grown through mergers and acquisitions.

One telco implemented governance over its enterprise product catalog. The operator had more than 20,000 Universal Service Order Codes (USOCs) in its legacy product catalogs. However, the operator decided to implement a new enterprise catalog with only 500 products that accounted for 99 percent of the revenues. Because of the new catalog, product management was able to reduce the time to introduce new products by 70 percent, which increased the revenues for the associated products by three to four percent.

Finally, the operator was able to shave four weeks from the training time for new customer service representatives on the simplified product catalog. This was important because the average tenure of a customer service representative was only 12 months.

Dylan Jones: You’ve talked about the importance of standardisation within telecoms, can you give an example of how have you personally helped companies benefit from this in the past?

Sunil Soares: I have worked with a number of telcos in North America, Europe, and Africa. My work has ranged from helping them establish their Data Governance Council, adopt a charter, identify critical data elements for a data quality scorecard, identify data stewards to support a master data initiative, and starting up a business glossary.

Standardisation is also an important aspect of data governance in terms of consistency of critical business definitions. For example, Average Revenue per User (ARPU) is a critical term that is relied upon by executive management to make business decisions. However, marketing, finance, and sales often have inconsistent definitions for ARPU. For example, finance might include employees while marketing might exclude employees from churn calculations.

The Data Governance program needs to drive to a consistent definition for critical business terms.

Dylan Jones: Finally, how would you recommend a telecoms company get started with Data Governance of their Big Data? What are some of the key takeaways you’ve discovered from your years of experience in the field?

Sunil Soares: Here is my Industry-Oriented approach to Data Governance that applies very nicely to telcos:

  1. Define business problems that are specific to your industry, job function, and company (e.g. customer-centricity for marketing in telcos)
  2. Prioritize data domains (e.g. customer data)
  3. Identify a handful of Critical Data Elements (e.g. “household” and “address” to support cross-sell and up-sell programs)
  4. Write a Data Governance Charter
  5. Create a Responsible-Accountable-Consulted-Informed (RACI) Matrix (e.g. Is Marketing accountable for Household data?)
  6. Quantify the financial benefits
  7. Define the Data Governance Organization including roles and responsibilities
  8. Establish a data quality scorecard starting with a small subset of KPI’s
  9. Stand-up a business glossary starting with a targeted number of key business terms (e.g. Does “household” include students who live away from home but are on their parents’ calling plan?)
  10. Align with the technical architecture