An Insight into the Business Intelligence Needs of a Telecoms Operator
Well then with this depth of information it should be fairly easy to work on and keep the telco customers happy, right? Well as much as I would want to believe that it should be fairly easy, it is not so. Today the telecom players face a huge set of challenges with Operational efficiencies, network stability, service offerings, revenue leakage, customer satisfaction and above all churn and Average Revenue Per User (ARPU) Management. And BI plays a very key role in helping to address many of these issues and problems.
The Telecom and Media market in the last few years has undergone a silent revolution in terms of network convergence i.e. using a single network to deliver voice, data and content services. Now you have a single service provider catering to your landline needs, Internet needs, mobility needs, Television and Entertainment needs. This is a revolution which has happened over the last half a decade and the entire world over. So when we talk about a Telco operator today, chances are that it would be providing most of these services to its end customers.The Business Intelligence and Reporting needs for a Telco could primarily be classified as A).Operational and B).Strategic or Predictive needs.
As far as the Operational BI needs are concerned we can see a good mix of needs between the both the OSS and the BSS stack. While the Network teams in the Telco might be keener on understanding the network usage and utilization, Quality of Service (QoS) KPIs, SLA adherence especially for the high valued corporate customers and above all the Inventory utilization. Networking inventory forecasting and utilization continues to be by far a very challenging subject for many Telco operators and although some packaged solutions do exist but a lot is still to be done. These metrics and KPIs are very critical and any Telco operator should have this pulse because for all the rich content and value added services that it might provide, relatively poor or inferior network coverage can pull it all down.
Also quite operational in nature are the needs of the various Product and CRM managers of Telecoms operators who wish to track the Order Processing, Service Request Processing lead times, success rates along with the call resolution satisfaction rates and efficiency scores. In order to keep your operational costs down it would make all the sense to be able to provide as many resolutions remotely without having to do a ‘truck roll’. Truck rolls or where a technician needs to visit a site in order to fix a complaint are usually an expensive affair and best avoided. There are various other similar parameters which a CRM manager would wish to monitor in order to ensure efficient and a functional support staff.
I would be doing injustice if I do not mention the biggest of them all and that is Revenue Leakage Management. It is possibly any telco player’s nightmare and a huge area to improve your bottom line. A study done a few years back, said that annual revenue loss due to revenue leakage for all major telco of the world put together runs into around 40-50 billion USD. The reasons for revenue leakage are many and occur at different junctures of the OSS stack and also at the inter-operator connect. Any telco would be looking for the need of a robust intelligence system to identify and address this issue.
While I believe that the Operational side of telecom BI are more or less figured out today and is more of the science, it this strategic or predictive BI needs which is the art and is often the more difficult to crack. In my interaction with various product managers of several telcos across continents, I have learnt that the top two things he/she grapples with are “Churn” and “ARPU”. ‘Lower Churn and increase ARPU and you are there’ - that is what the Telecoms CEO tells his/her product managers. Easier said than done! Especially in the Western European market which is possibly the most matured, advanced and above all a completely saturated segment, Telcos are facing the most daunting task of holding on to their existing customers, waging a price war and at the same time expecting to improve the top line revenues.
BI to the rescue? Well yes and no.
Calculating and Predicting churn is done by different operators in their own little way and also there are no fixed set of rules to either report or predict the churn. Although calculating the churn can be fairly standardized, predicting it is done based on different algorithms and even use of artificial intelligence. There are some standard packages available in the market which helps you identify and work on churn predictions. But at the end of the day these packages or systems are all sort of a data mining tool which sits on top an Enterprise Datawarehouse where a churn probability score is provided to all the subscribers or customers based on how the telco perceives end user behavior. The churn score is calculated based on social and behavioral algorithms. For example if a subscriber reduces his calls by more than 60% in one particular month his probability of churning increases. Or for example if the subscriber makes most of his calls to the competition’s network he is as well more likely to churn. There could be many more such parameters based on which the churn model works and there are numerous white papers and thesis which provide inputs on churn prediction.
ARPU on the other hand is again stimulated by customer’s call and products usage. Product managers typically study the usage patterns (based on various analytical exercises that get carried out) of the high valued customers and design specific marketing campaigns to stimulate usage.
Although there are many algorithms and packages to help the Telecom players to manage these two “devils” but it being an art, has still not been perfected and with changing times and human behavior needs to be revisited and remodeled constantly. That is possibly where the beauty lies!
