Systems integration has become a major challenge for pay TV operators as they embrace IP infrastructures and online delivery, with a requirement to become as agile as Internet companies while containing costs and dealing with legacy. For many the choice of integrator has become more critical even than that of the infrastructure’s core components, because getting it wrong can derail the whole enterprise and risk losing valuable ground at a time when new entrants are arriving and competitors are innovating faster than ever before in TV’s history. Against this background I have just co-written with Ben Schwarz an eBook focusing on SI in pay TV and drilling down into the core issues through interviews with leading operators, analysts and of course integrators themselves. What emerged was a clear picture of best practice and priorities both in selecting an SI and then managing the subsequent relationship.
Some of the key takeaways may sound surprising at first sight, with a principle finding being that innovation does not equate with novelty and that for established pay TV operators it absolutely must carry legacy with it. This means that while the UI and features like bookmarking and recommendation should keep pace with OTT providers like Netflix, as much of the added value as possible must be delivered to legacy set top boxes. It is impossible to swap boxes at a stroke among a subscriber population of several million and legacy will always be a fact of life that must be embraced by SI.
Mark Perry, CEO of Oregan Networks, provider of embedded middleware for TV delivery, expressed this well during our interview, pointing out that sound legacy hardware management was key to successful next generation transition projects. Operators, he said, must limit service disruption and deliver OTT along with non-linear content to existing platforms in the field.
There is a dichotomy here in that while operators naturally want to hold onto legacy equipment as long as possible to postpone upgrades and contain capex, they also want the new functionality to be available quickly. As Telefonica’s IPTV Senior Project Manager Jose Maria Bravo pointed out, cloud storage can help here by extending benefits previously only available on advanced set top legacy boxes with no hard drive.
Then Perry noted that users’ legacy boxes must benefit from the same agility as those with the latest kit. This can be achieved by installing new client middleware on STBs from a remote server, with no disruption to the service, after which features can be turned on quickly and automatically. Inevitably though it takes more effort to develop middleware enabling such agility on old set tops and this may be a factor in choice of SI partner, which should have the capabilities for the task.
Another point immediately follows, which is that as speed to market for new services, features and apps becomes a critical differentiator, development time and the friction imposed by testing must be reduced drastically. For this reason DevOps has become a mantra for pay TV, shortening development and testing time by bringing development and operations closer together in both technical and human terms, increasingly with the same team handling both.
At the same time the importance of virtualization became clear though our interviews, with the crucial point being that it lets development teams adapt how services are delivered without needing to change hardware. The location of nonlinear TV services is a good example, where such flexibility makes it possible to deliver the best user experience. Simple time shifting like pausing a live stream for a few minutes can be done locally on the set top, while longer recordings can be stored in the cloud for greater efficiency. Rather than the cloud taking over CPE-rendered services altogether, agile operators are talking more of the “cloud-enablement” of CPE, aided by virtualization now extending right into the home as far as the set top or gateway.
In association with DevOps and virtualization, pay TV development teams are adopting a micro services approach, shifting away from code-heavy monolithic applications to smaller, self-contained processes that can be introduced with minimal disruption and low risk. This enables the rapid deployment of new features on a continuous almost drip feed basis as they come, reducing time to market from months to days.
One further requirement coming out of our research was also something of a contradiction. While on the one hand operators are seeking a kind of one stop shop where there is just a single butt to kick when things go wrong they also want to exploit best of breed, in some cases just to differentiate their service from the competition by offering some novelties associated with a given component. They don’t want to have exactly the same infrastructure as their competitors. For this reason they are looking for a component integrator which can enable best of breed while insulating them from the complexities involved. So as Pav Kudlac, Service Line Director for Video Systems Integration, Nokia, put it during our interview, it is vital to have one throat to choke, one integrator who is responsible for all delivery even if contractual arrangements are with individual suppliers.
Ultimately though the operator should stay in control and ensure that the integrator does not call the shots. This means putting contractual arrangements in place such that the integrator has the incentive to deliver a great service cost effectively. This may all sound like a lot to think about but, as we found, if the right SI is appointed with the right contract, the operator has a strong competitive foundation for going forward in an uncertain and rapidly changing media landscape.