Accelerate innovation and harness Big Data with the Cloud

    Share on

    There are different definitions of the Cloud and what it means for Pay TV. Some vendors and operators associate the Cloud with unifying and upgrading existing headends to enable multiscreen service delivery. For Cisco and NDS, this as an important first step but there is more to it than that. Solutions based on real cloud computing infrastructures can centralize middleware, metadata and UI logic to reduce complexity of clients, especially the STB, leading to faster innovation and cost savings. Another characteristic of cloud computing, elastic provisioning, means Pay TV operators could also pay cloud vendors for functions like storage or transcoding based on usage instead of maintaining hardware infrastructure that could be under-utilized outside peak hours. It is claimed that this will help reduce both CAPEX and OPEX.

    It is generally acknowledged that the Pay TV industry needs to innovate faster, partly because of the competitive pressure from OTT challengers and partly because of the compelling nature of some CE devices (and of course, these two trends go hand-in-hand). But OTT is also a big opportunity if operators can master the same web based technologies and exploit centralized cloud management to become more agile and faster to market, and stake out their part of the new multiscreen experience.

    According to Kip Compton, VP & CTO Video and Collaboration Group at Cisco, a great benefit of cloud based architectures lies in being inherently more future-proof. He cites the Web UI for Gmail that adapts automatically to different screen sizes, so when new devices arrive, cloud software will simply adjust the UI to fit the relevant display technology. The ability to fix and update software quickly is also a compelling advantage of cloud in terms of being able to innovate and respond to the market quickly.
    Compton sees a key role in the new operating model often called “Devops” already adopted by Google, Facebook and Netflix.

    “These service providers can have a software fix operational within 12 hours. By contrast, the inherent complexity of networks and device management mean that the most agile of traditional Pay TV operators only get a new software version out every six months. That’s one reason why the cloud gives operators the ability to innovate far more quickly.”

    Despite the clear advantages, broadcasters’ development teams are not yet swayed by pure technology arguments, according to Nick Thexton, Vice President and Chief Technology Officer, Service Provider Video Technology Group at Cisco. The business leaders are often ahead here, as a cloud approach to TV means that new services can be rolled out in weeks or even days rather than months. A big part of this is the ability to perform so called A/B Testing to assess which of several different versions of a service improvement will generate the best response from customers, relying also on operators’ big data capabilities.

    “A/B Testing is where an operator picks out a small sample of users, say 100, and gives them access to the sample change,” says Thexton. “The operator can then measure in real-time if usage improves or not. That’s what Netflix have been doing to hone in on an optimal buy rate. Most Pay TV operators are not quite there yet, but this is clearly on the horizon.” Thexton goes on to point out that an inherently web-based architecture provides better integration with external systems like micropayment or personalized social media, further speeding up innovation. One major benefit for operators, in terms of speed to market, is in being able to apply innovations and service changes almost instantaneously across their whole target device population. This is essential for multiscreen services, ensuring for example that a cloud-based application can retain the same functional logic, just adapting the UI to different devices.

    Looking further ahead, the cloud will be the key enabler for Big Data to deliver on its promise to improve live TV services. Letting unmanaged devices directly access a cloud based server will bring a new wealth of cross-device usage data that will help operators better understand user behaviour. Apple clearly understood this in its transition to iCloud, initially forcing all devices to be tethered but later lifting this restriction. Nigel Sanctuary, VP for Cloud Propositions at Kognitio Cloud, states the goal as being to bring new and large volumes of data to bear to provide a crystal clear granular view of consumer behaviour. “For years the Analytics industry has been telling operators what users have been doing and extrapolating what they might do in the future. Now, thanks to the cloud, real-time responses, or at least responses within an active user session, have become possible so we will be able to tell operators what users are doing in the present.”

    Sanctuary says recommendation and personalised advertising can already be achieved now as all the technology is here. There is, however, still a challenge in the scaling of any such real-time solutions to user bases measured in multiple millions. “Pay TV’s challenge is to bring the scalable real-time network and usage data into the analytics environment, but that is part of what the cloud will enable.”

    The potential of cloud based recommendation to help steer subscribers through the increasingly large range of content available is of great interest to operators such as Viasat. “On the Internet, unlike traditional broadcasting, everything is measured,” says Martin Lewerth, Executive Vice President of Pay TV and Technology at Viasat, the broadcaster and satellite distributor owned by Swedish media conglomerate Modern Times Group. “We know that with thousands of TVOD (Transactional VOD), SVOD (Subscription VOD) and catch-up titles available to the consumer, the decision making of what to view is becoming increasingly difficult. Personal recommendations, based on online viewing data, will be instrumental to simplify usage of new cloud-based services when the width and depth of content availability is increasing.”

    Increasingly, operators will want to perform multiple queries across large numbers of sessions and massive data volumes in order to make decisions that will provide vital competitive advantage, and this will require huge processing capacity. This effectively can only be done in the cloud because the hardware resources have to be in one place that also has access to all the relevant data. “Parallel processing is extremely effective in the cloud because it provides not just storage but also the CPU, and can detect patterns in your audience measurement data, and get the information in real-time, to answer questions like what people are watching now,” Thexton explains.

    You can read more about the potential for cloud computing to help Pay TV operators improve their services and reduce costs in the Videonet report â€˜Cloud computing at the service of Pay TV’.

    More about the report:
    This report is a look at what the ‘cloud’ means for Pay TV operators and how they can harness cloud technologies to accelerate innovation, reduce operational costs and explore new opportunities including those surrounding ‘big data’. This report argues that unifying existing headends to enable multiscreen services is only the first, albeit important, step into the cloud. Solutions based on real cloud computing infrastructures can centralize middleware, metadata and UI logic to reduce complexity of clients, especially the STB. Elastic provisioning means Pay TV operators can pay cloud vendors based on usage instead of maintaining under-utilized hardware infrastructure. This report includes input from Viasat, Comcast, Liberty Global Europe, Rogers Communications, Cisco, Kognitio Cloud, Amazon Web Services, Analysys Mason, Arthur D. Little, and IHS Screen Digest, among others. Read report here.


    Share on