RPA for contextual intelligence in the telecom industry
An intelligent, non-disruptive and scalable robotic process automation (RPA) system can present many advantages — meaningful data collation, smart insights, saving in time and money, and most of all, competitive advantage.
Few will debate that telecom is an industry majorly disrupted in the past decade. Economies of scale of manufacturing have made mobile phones more accessible to the average Joe on the street, while competition between service providers has led to price wars that are won on numbers over margins.
Growth, but at a price
Telecom companies, for their part, have leveraged business process reengineering in every way possible — from integrating themselves across the entire value chain to outsourcing every single activity they can afford to.
This growth, across the length and breadth of telecom organizations, has resulted in systems that are disparate and non-communicative. Many of them were built as the system evolved and the customer base grew, but not all of them were built to tackle the complexities that could arise out of the scale they are growing to. For instance, an in-house CRM may be efficient handling a million customers’ details, but a multi-national telecom major will need multiple CRM systems to cater to bigger volumes of data, not to mention additional geographical, legal, regulatory and cultural requirements.
The cost of human resources
A recent survey of telecom companies revealed that automation on a large scale, backed by artificial intelligence and machine learning, has not made much of an impact so far. This also explains why, for such firms, the cost of human resources for data collation and processing is high enough to significantly dent their bottom line.
Sample this: in a typical telecom organization, repairs are reactive. This means that something has to go wrong and be reported before the company will commit resources to fix the problem. This has multiple costs. For one, there is the economic cost of downtime in all the billable calls that could not be made. And the loss of customer goodwill (both corporate and individual). For telecom companies that offer additional services such as the internet, conferencing, maintenance, etc., there might even be punitive damages built into their contracts for non-fulfilment.
The technologies involved are themselves so complicated that humans simply cannot be relied upon to connect correlations where none are readily apparent. Even the smartest technicians need to have non-essential data filtered out before they can detect anomalies in the system. As human beings, they also need time off to refresh themselves lest they miss a warning due to fatigue. This commitment of resources cannot be avoided in the absence of an alternative, even if the system is extremely stable. In fact, such redundancies are essential when the system breaks down.
The case for RPA in Telecom
Robotic Process Automation employs a wide array of bots that can scrape websites and display screens, databases, ERP systems, excel sheets, printed text, etc. and generate useful data from all of them. A good RPA can not only collect data but also collate it to generate insights that are useful to key stakeholders within the organization. Organizations without RPAs tend to do this the old-fashioned way by committing manual effort — hundreds of thousands of hours a year. The cost-savings from moving to an RPA-assisted system itself will pay for the investment in RPA within a few years.
RPAs allow managers to devote more time to analysing data than in putting it together. RPAs can also be set up so that they send automated reports at fixed frequencies (hourly, daily, weekly, etc.) to the concerned managers, which means decision cycles are significantly shortened. RPAs can function in the continuum that stretches from customer-specific granular data to organization-wide macro data. It can identify upselling/cross-selling opportunities to increase the lifetime value of a customer.
RPAs can also scrape third-party channels like newspapers and industry reports to gather additional information on assets and competition. For instance, if inclement weather is expected in a region, the transmission nodes servicing it can be allocated additional backups. If a lot of repeated calls — which indicate a higher incidence of dropped calls — are made in an area, an investigation can be underway even before the first complaint comes in. Sentiment analysis on social media and subscription reports from third-party channels can throw more light on competitors’ strategies and how they might be dealt with.
One of the most significant advantages of an RPA technology is that it is non-disruptive. Where a typical ERP will require extensive testing and possible downtime (due to incompatibility with existing frameworks, fixes, cascading faults, etc.), an RPA sits in front of the system, reading off but never really affecting existing databases. It is the RPA itself that is modified: it is trained to learn from the systems at the clients’ end and will need to know how to access these systems after all.
Good RPAs also tend to be scalable in terms of both scope and effort. Organizations can choose to deploy more bots on a particular module to improve the speed/quality of insights, or spread them out to cover more areas of operation or find a sweet spot between the two extremes that addresses the company’s requirements best. The technical barriers to effecting such on-the-fly alterations are low, which keeps the company’s costs low as well.
Robotic Process Automation is clearly the next big wave of disruption that will affect the telecom industry, making a crucial difference where it really counts — in every cost center, every second of the day. It might just be what separates the winners from those who fall by the wayside.
