In 
The Telecom infrastructure space has a lot in common with other diversified infrastructure providers, for example, support power or water distribution companies. The challenges faced by these sectors are very similar. These are
- a geographically dispersed network of points of presence across different environments and climatic conditions
- multiple assets deployed at these points of presence requiring operations and maintenance support
- a large and distributed field force to handle operations and maintenance of the equipment placed at these points including preventive and corrective maintenance
- diverse project teams to handle installation and commissioning/decommissioning of assets at these points of presence based on changes to the network
In addition, these companies need to build and maintain applications that help to –
- continuously or periodically monitor these points of presence for service availability and performance
- ensure that all company assets at these points are in good working condition delivered in the required functionality
- ensure the physical security of the assets and that they do not pose any safety risks to the surroundings in which they are placed, and that they are not tampered with by unauthorized personnel
- manage their diverse partner networks to ensure that services are being delivered as per their agreements and that the work that is expected to be done is actually being done in the field as per the specifications of the work
From a commercial perspective, companies need to accurately measure the services they provide, and service levels maintained and a robust mechanism to accurately (and promptly!) bill their customers for services received, as well as robust financial systems to manage their costs, receivables, and payables.
Needless to say, all of these transactions need to be carried out with full observation of all legal and statutory requirements ensuring the highest standards of corporate governance.
From a transactional point of view, therefore, such companies have a huge scale in terms of the number of transactions and need processes and systems that are capable of handling both high volume and velocity with speed and accuracy. Data volumes also tend to increase exponentially with new and innovative products and services being introduced.
Beyond the transactional space comes the entire gamut of the analytical space i.e. using the data generated from transactional systems to analyze process adherence for the company with a view to continuous improvement as well as compliance reporting. This involves the identification and management of exceptions, the establishment of dashboards and workflows with approval/escalation hierarchies to ensure appropriate controls in managing these transactions. Some useful metrics in asset and service-intensive businesses also include price-performance-quality, SLA adherence, bonuses and penalties, and the establishment of standards and norms across different geographies and environments.
There are also statutory and business requirements for time series data over multiple years to be archived, processed, and analyzed to provide inputs to commercial, sourcing, and partner strategies.
Advances in equipment technology across industries have given rise to “smart” equipment that continuously reports their status and progress, providing both diagnostic and performance data that must be continuously consumed and processed since corrective action needs to be taken in near real-time to avoid waste or other risks in case of malfunctions.
All of the above indicates that these businesses are extremely data-intensive and that CIOs in these businesses need to pay special attention to the challenges of collecting, standardizing, maintaining, and analyzing these data sets or risk losing out on their competitive advantages of scale and momentum. In addition, CIOs in these businesses need to understand the value of their data and treat it as a precious asset, i.e., build in the levels of security and access control to ensure this data does not get exfiltrated into the public space due to the commercial and (in some cases!) even security implications of this data being available to the larger public.
At Indus Towers, we have, over the last several years, integrated much of our available data into a state-of-the-art data warehouse, designed in such a way as to cover every data source in the organization and organized to facilitate combining of this data to make relevant analyses (and identify exceptions) – something that is not possible in siloed traditional business applications like ERP and CRM platforms. By intelligently combining data into relevant business areas, we can provide a layer of abstraction to our user community to be able to analyze data across different filters and parameters, providing regional management a day-to-day reporting for operations support while also providing corporate functions and senior leadership with strategic trends in usage and business parameters.
The mechanics of “feeding the beast” and ensuring appropriate capacity and performance for this system end up being a top priority for the Information Technology teams at these organizations. There is an obvious tradeoff between multiple times a day refreshes vs. (say) daily end-of-business-day replication and synchronization of data and processing of dashboards etc. The value of measuring data in the movement has to be keenly weighed (and the use cases and benefits clearly understood) before creating the massive infrastructure and visualization involved in real-time or streaming data situations. I believe that most business cases permit overnight refreshes so as to ensure that systems can be checked for consistency and that we do not place excessive load on operational systems at the same time that users of these systems are trying to get their day-to-day jobs done!
On the visualization front, it becomes important for IT Teams to deliver to their citizen analysts -a good documentation and understanding of the underlying data structures, their metadata, and the business association of that metadata with the parameters that they seek to measure. Ideally (as we have done in Indus) providing training and a self-service interface to build your reports, with access to the data across levels helps us in managing transactional reconciliations as well as allowing for free-flow strategic analyses.
From a technology landscape, there are very different stacks of solutions that are appropriate for differing levels that data size and complexity, the sophistication of citizens and other analysts, and archiving/backup and restore efficiencies which may be required by the business or industry. It is necessary for Information Technology teams to understand the inflection points at which the current components of their technology solutions become inefficient or even dysfunctional and create their roadmap with their milestones that indicate the need to augment, upgrade and sometimes replace their technology stack with bigger and better solutions that are more appropriate to the size, complexity and, scale of their transactional, tactical and strategic analytics needs.
Indus’ own journey from disparate siloed applications and citizen analysts combining data in Excel has been long and arduous but totally worth the investment. From one source of data, we are able to provide business-as-usual operational data that drives the day-to-day activity as well as the periodic exception analysis and medium-term reporting that provides inputs for tactical and strategic decision-making. Significant investments have gone into ETL and data organization/storage as well as in front-end visualization and we are now in an optimization phase, removing unwanted data sets and eliminating duplications across functional areas while continuing to find the blind spots of data that had not made it to our radar so far. Some advanced use cases of “closing the loop” i.e., using decision support data to drive transactional corrections or actions without human involvement through RPA or scripting are giving us significant benefits in terms of TAT and productivity.
It is a testament to the quality of thought put in while setting up this decision support system that over a short period of two to three years, this has become one of the core platforms of the organization and ranks on par with our ERP, CRM, or field force management systems in terms of utilization or strategic importance. Over time a lot of our strategic measures have been defined and their logic sharpened till we now have a repository not just of our data but the basic algorithms of our business in terms of how we operate. While data in and of itself has tremendous strategic value, that value cannot be unlocked without a robust and flexible way of collecting, securing, storing, and expressing the data to drive operational, tactical, and strategic decision making and reporting.
Future directions are to further optimize the data collected and stored as well as develop more unique use cases, particularly to feed the shift from reactive to responsive (which has largely happened) and now from responsive to predictive analytics.
The author is CIO, Indus Towers Limited
Add new comment