Monetising Big Data in Telecoms
Author: Andrew Tan CIO
Date: 22nd November 2016
Categories: Technology, Data, Financial, Banking, Telecoms, Revenue Assurance, Fraud protection, Risk Protection, Billing Systems, Billing & Charging, Neural Technologies, Big Data, Analytics, CAPEX, OPEX
These days, when you talk to Communication Service Providers and technology partners in the communications space, the term “Big Data” is always quick at hand. There is no conference that you can attend where Big Data doesn’t play a major role, and everyone seems to be interested in “what are you doing in Big Data?” or “what is your Big Data story?”
Most system landscapes at CSPs are not designed to gather all the information which is available, and even if they are, they may not be ready to share the data. Add to this the task of setting up a functioning Big Data analytics environment and it becomes clear that the initial and ongoing investment in this type of technology is big by any standard.
To make matters worse, even if you had a Big Data strategy in place you are unlikely to have qualified staff at hand to get the most out of your investment. The market for skilled Data Engineers is drained and will be for some time.
However, with CSPs battling to survive in largely saturated markets, there is little choice but to march into this new frontier.
What do you need to be successful in Big Data?
A recent study by Neural Technologies revealed that 50% of CSPs said that they already use Big Data tools to visually analyse and understand risks, while all others said that they don’t know or have no plans at the moment.
I personally don’t think there’s a one-size-fits-all strategy. What’s important is that you get started somewhere. Even the longest journey starts with a single step, which is likely to include: a plan, a data-friendly systems environment and the right team!
There are various ways in which the data can be turned into profit:
- Understanding customer behavior and deriving new revenue streams in the form of innovative product offerings
- Selling information about customer behavior to third parties. The more precise your subscriber data, the easier it is to carve out new niche products and more targeted marketing approaches
- Improving the customer experience through better information automatically generates more profit as it commonly increases usage and reduces churn
- Exploit the ability to differentiate from the competition and reduce overall CAPEX and OPEX by better understanding where to invest
The impact of Big Data on billing
The insight you can gain about your customers seems almost endless. Companies like Google, Amazon and Facebook do this on a continuous basis and in real-time. It’s time for the CSPs to do some catching up.
CSPs are only scratching the surface of what can be achieved. Especially because the data is already there, customers are already willingly leaving all kinds of traces. CSPs know where their customers live, who they communicate with, which games they play, which websites they visit, what their daily routine is and also where they travel. When you think about it, there should be far more information available to CSPs than to Google or Facebook.
What’s the impact of all of this on billing systems? Firstly, as data volumes increase - the load on the billing system increases. Suggest to any CSP that their billing system had to process twice the current amount of data within two years’ and you would see some faces go pale.
Most CSPs run legacy billing systems which in turn rely on legacy architectures and legacy data stores which just aren’t designed for gigantic amounts of data.
Billing systems are good at valuing (rating) data based on specific parameters. When you think of this in the context of Big Data, we see that there is the capability to ease analytics of very large data sets by pre-valuing specific pieces of data. Do this in real-time as the data arrives and you can derive instant pre-assessed views on your subscribers and possibly derive real-time actions (e.g. offers) from them. It’s like an instant determination about which data is important and which is not so valuable. I predict that in the near future, we will see more intelligent ratings based on machine learning and anomaly detection algorithms in telecoms systems.
We are really just at the beginning, expect to see big leaps of development within the next five years. Much of this is in the early experimental stages, but just as traditional data warehouses did their fair share in improving service, cost and margins, Big Data will inevitably take us to the next level.
What is your Big Data strategy?