Rajeev Tankha is the vice president of revenue and customer management at Amdocs. Here, he tells George Malim how policy and charging have converged and the focus of communications service providers (CSPs) has turned to using the two to enable charging for quality of service, greater personalisation and an analytics-based feedback loop for CSPs to assess the effects of network and commercial policies on customer satisfaction and business performance.
VanillaPlus: LTE networks enable greater capacity, speed and quality of service (QoS) control. How critical has policy and charging become in terms of enabling CSPs to differentiate and charge for services based on QoS?
Rajeev Tankha: In the new digital landscape, CSPs can create a more dynamic quality of experience that can be monetised in different ways based on customer profiles and service requirements. Enterprise customers, for example, have different QoS needs at different times of the day – and are willing to pay for those guarantees. For example, a payroll department might need guaranteed QoS for a short time while records are being processed and transferred. CSPs need to be able to support that service level agreement (SLA) requirement and charge differently for it. The policy and charging rules function (PCRF) can manage the dynamic QoS change but it has to work hand in glove with the online charging system to capture the revenue associated with it.
On the consumer side, a good example is video services. We know that a high percentage of LTE traffic comes from video, and usage is expected to rise exponentially as more services get rolled out over the next few years. If CSPs can monetise the value of HD service quality for a user who consumes a lot of video and finds that increased QoS valuable, they can generate higher ARPU and engage customers with more personalised offers. Again, that requires tight integration between policy and charging systems.
VP: Personalisation enabled by policy looks to be a great way for CSPs to tailor higher value propositions to individual customers or sets of customers. Do you see personalisation as a way forward for CSPs in their monetisation efforts?
RT: People tend to think of personalisation as relating to consumer users or businesses but the reality is we’re a hybrid of both, especially as the enterprise becomes increasingly mobile. Users’ requirements are different at work than at home, so if CSPs can design policies according to each persona and charge for the different service profiles, the potential for monetisation increases significantly. Policy and charging control (PCC) systems have to be tightly synched to apply policies in a contextual manner and deliver the quality of experience that will increase the value of the service and squeeze more revenue from the use of the network.
VP: So what are some of the requirements for delivering contextual services and greater personalisation?
RT: Sharing the same subscriber profile repository (SPR) database enables the PCRF, instead of just counting upload, download, bytes and durations, to use business logic that enables CSPs to perform actions such as charging and network resource allocation or to take decisions and apply new policies. In addition, the integrated system can be used to count monetary values, such as to check a customer’s remaining balance before authorising a service.
A tightly integrated policy and charging control architecture and its utilisation of a Sy interface – standard Sy and Sy plus – enriches Sy in areas where the standard has not been defined. This Sy interface adds value because it enables business insights from convergent charging via the common SPR to be exposed to the policy controller. A good example of this would be monetisation of a one-time uplift in data speed for a user via a turbo boost button.
We’ve architected the Amdocs Policy Controller and Convergent Charging system with a shared customer data repository, standard Sy support and Sy plus integration so that our customers can reduce the design time to create innovative new services and deliver tight synchronisation of policy and charging events at runtime.
VP: What about the impact of guaranteed service levels on a relatively new service such as voice over LTE (VoLTE)?
RT: High definition voice quality can’t always be guaranteed – for example, in the case of network congestion or on devices that don’t have HD support. CSPs need to have the capability to charge differently for guaranteed versus best effort QoS to provide value and ensure retention. From an opex perspective, VoLTE signalling generates a tenfold increase in per-subscriber costs to deliver HD voice. These are costs that can’t be mitigated by added revenues so it’s critical to have low total cost of ownership (TCO) policy systems that are highly efficient and designed for optimum performance on commercial-off-the-shelf (COTS) hardware to keep the costs of delivery in line with revenues.
VP: You spoke earlier about enterprise services being fertile ground for CSPs to seek to monetise their networks more effectively. How do you see enterprise propositions maturing?
RT: Let’s say a movie studio customer wants extremely high quality bandwidth for a fixed time to send digital prints of a new movie to pre-production. That demand is for a specific length of time and normally at the same time each day during production of the movie. The movie studio is willing to pay four or five times the normal cost to get those large files sent at the specific time.
That’s just one example from what we’re seeing in the field that shows how much service delivery has evolved. Dynamic policy control is the underlying foundation for enabling the service and allowing the CSP to guarantee the customer the performance they need when it’s required. This is a strong monetisation opportunity but to execute on it effectively, CSPs have to provide the enhanced service and prove that it has been delivered at the quality customers expect.
VP: Services such as the one you describe above are complex to set up and bring to market. How can CSPs accelerate their innovation to make these types of proposition real in a timeframe that is acceptable to customers?
RT: When you’re defining a service, you have to look across two broad areas. One is setting a commercial policy, essentially how to price a service, and the other is setting a network policy. Usually different experts are involved to set each and building commercial policies takes time. At Amdocs, we’re helping CSPs address this challenge by creating a common design time environment that delivers a common tool across all systems to create and design capabilities once and then launch the service.
Another aspect to rapid innovation is virtualisation. Our policy and charging control system can run in a virtual environment which means that you can instantiate another perfect copy of your production environment in a matter of minutes. That’s fundamental to being able to quickly onboard enterprise customers and test new capabilities and services without interrupting the operations of the live network. That takes weeks and a lot of development and testing resources to do with physical big iron.
Here, Amdocs provides the advantages of our common design time approach and our ability to use virtualisation to customise data plans per enterprise and even department, and instantiate those data plans in less time and at a lower cost than traditional networks.
VP: Policy isn’t just about setting up new services, it can be utilised to provide insights into business performance. How do you see analytics of policy helping CSPs run their businesses efficiently?
RT: One of the greatest concerns CSPs have is customer churn or the propensity of customers to churn. That usually is dependent on network quality and CTOs typically look to network policy analytics to gain insight into issues that could be affecting their customers’ quality of experience. Measuring the impact of network policies can be highly valuable for setting customer retention strategies.
Examples of areas where policy analytics can help drive retention and revenue are in HD service delivery and fair usage impact. Policy analytics can provide measures of the level of high definition versus standard definition service quality for VoLTE when 20% of your users are getting downgraded to standard definition or determine the effect of strict fair usage policies that are resulting in aggressive throttling.
In addition, analytics can be used to improve monetisation strategies. For example, analytics can be applied to explore why a new service launch doesn’t deliver as expected or why a new service isn’t being taken up or resulting in subscription renewals.
In summary, we can see how the move to the all-IP network is creating new levers for CSPs to increase the value, uptake and revenue from data services. CSPs can use dynamic QoS and speed for high-value contextual offers, deliver greater personalisation for different subscriber personas, and use virtualisation and analytics to increase service agility and close the loop on network policies and the subscriber experience. We’ve architected the Amdocs Policy Controller and Convergent Charging system to support these new monetisation and retention strategies. At the foundation of our approach is a tight integration of the Sy layer to support a common runtime and customer database, which is being used today by CSPs to drive data plan innovation with millions of subscribers and operations working in real-time.
Comment on this article below or via Twitter: @ VanillaPlusMag OR @jcvplus
