Many communication service providers (CSPs) – from telecom operators to satellite and cable providers – have long operated in a world where they were the only specialists in providing their particular services. But times have changed and they now have to compete with other types of providers.
Whereas we used to call and text people, we now often choose to Skype or WhatsApp instead; we may have paid up to £40 pcm to receive a vast array of film channels – often including a lot of films that we’d never watch – but today we can pay £9 pcm for Netflix where we can stream a one-off film or TV programme on demand, or get Amazon Video as part of our Prime package. So CSPs can’t continue to depend on their traditional sources of income for profit.
This puts CSPs under constant pressure, not only from OTTs (Over-The-Top Content Providers) such as Facebook, WhatsApp and Google, but also from regulators, such as the EU Commission and from cloud service providers such as Amazon and Microsoft.
But CSPs still have unique assets in the shape of their subscribers and network which they have been building up for decades and which they can make use of to evolve their business model.
It’s certainly true that CSPs are facing heavy competition from OTTs, but changing an operator is still far more difficult than changing a chat app. Nowadays, people may install many different apps on their phones, but acquisition of users is easier than retention. A bad registration experience, privacy concerns or too many annoying ‘push’ notifications and people will uninstall one app and start using another one quite quickly. Changing an operator, on the other hand, is not quite so easy. So CSPs are – still – in a good position to make use of their subscriber database.
For companies who come from the world of computing, it may look as though building a network is just a question of putting together a few racks and connecting them via Ethernet cables.
But those companies – companies such as Google Fiber – have already failed in their plan to provide connectivity to a broader audience. At its start, Fiber was lauded as one of the most exciting disruptions ever for the ISP market, but it has yet to deliver on its promise. A combination of poor marketing, low price, lack of partnerships and inconsistent service has failed to create a meaningful subscriber database.
Other issues such as logistics, field services, maintenance, legal issues, customer service and an omnichannel option, are just few of the additional challenges which computing companies may find new and daunting. But for CSPs, these are core business activities which have been designed and optimised over many years.
The question is how can CSPs use these clear advantages of a subscriber database and network to keep, and then expand, their position in the ICT market?
The Three Value Dimensions for Big Data
Over the last few years, three key areas have been under discussion, but they have often been uncoordinated and, as a result, unsuccessful. These three are how to:
• improve operational efficiency
• improve customer experience and loyalty
• find new business models
There are plenty of opportunities within each of these categories, but the key prerequisite for CSPs is to ask themselves (and then answer) these four questions:
• what is their core business (mission)?
• how are they delivering it (business processes)?
• what is the feedback from the market (customer experience)?
• what are the new opportunities arising from new technology (new business models)?
So simple and yet, so difficult!
The first step to realising these goals is to integrate technology with the business process. Without a methodology to monitor the bottlenecks within the organisation efficiently, CSPs will not be able to improve their operational efficiency, no matter how good the tools or the network technologies that they buy.
Another issue is that of making experienced staff redundant in order to cut costs and replace them with automation. But doing this may result in the loss of know-how in network planning and design and operations and leave the company’s remaining staff fearful and demoralised. Many experienced engineers will leave due to this pressure from this sort of cost cutting, rather than the pressure from increasing operational efficiency itself.
So, this is the dilemma that CSPs face in staying ahead of the competition: CSPs need to automate a number of their operations to replace the experienced people that they’ve lost, but the automation of operations requires experienced people to teach the system how to work – squaring the circle.
Fortunately for CSPs, there are a number of people who have great experience in network operations (either on the technology or on the business process side) who come from CSP support companies that are also delivering OSS and BSS solutions. So, there is a chance to fill the gap left by the missing experts with intelligent management platforms supported by machine learning algorithms, set up by experts with a great understanding of CSP specificity and which can successfully realise the decision cycle in an automated way.
Closed, Open or Semi-Closed Loop Decision Making
The aim is to reach a closed-loop model of decision making but the truth is that in a CSP environment which is extremely complex (either from a business or technological perspective), reaching the closed-loop on a global level might be challenging and, even if possible, very expensive.
In some areas, it might be reachable, for example, when setting up a self-optimising radio access network (RAN), but as a rule, we should assume that the semi-closed loop approach with a human being as a supervisor over and above a machine generated recommendation is a much more realistic approach. A promising approach is to monitor the actions performed in different systems by experienced engineers who are responsible for solving tough problems in the network. Doing so we can teach the system who will be able to move to the next level in automation, or at least will be able to provide recommendations to less experienced operators in the future. This is a new opening for a promise of building know-how database which was rarely achieved in an efficient way due to lack of self-discipline within the organization.
In part II of this article, we will look at the issue of data.