Chris Wood explains how to create an OTT infrastructure that provides effective direct-to-consumer capabilities and prevents strange things ruining the user experience.
Spoiler alert! In the cult Netflix series, Stranger Things, Jim Hopper’s decision to take the direct approach to solving Hawkins’ mysteries sees him swallowed up by a network of vines in the ‘Upside Down’. Oblivious to the dangers that lurked beneath, he almost pays a heavy price.
It’s a similar plotline for service providers and brands, whose efforts to establish direct-to-consumer capabilities typically rush to the endgame of the delivery network – only to get entangled in the reeds of technical challenges that should have been addressed earlier in the process. It’s a costly, Upside Down approach. Because if, like Jim Hopper, you don’t get the groundwork right, stranger things happen that can ruin the user experience and cost you viewers. That’s the ultimate spoiler. But it can easily be avoided.
The Demogorgon of direct delivery
Building the optimal platform to deliver Over-the-Top (OTT) content direct to the consumer isn’t easy. But, as the appetite for on-demand streaming increases, making the transition has become a high priority. Stirred by the growth of cloud-based disruptors like Netflix and Amazon, traditional broadcasters are launching their own direct-to-consumer platforms while niche channels and service providers are gearing up to join them. The problem is the move from traditional to IT-based infrastructure is a journey to a parallel dimension where stranger things happen. OTT delivery takes operators out of their comfort zones and requires them to acquire unfamiliar skillsets to create a seamless experience.
Delivering video content over the internet is fraught with peril that ultimately manifests itself in the customer experience. Viewers expect a great user interface, a stable experience and the ability to watch what they want whenever they want it, on any device. To be commercially successful, providers need to deliver high-quality streams on a global scale, where every region – and every device – presents its own peculiarities. There’s an awful lot to get right. What’s more, with audiences fickle and unforgiving, you need to get it all right, at speed. Time lags or choppy video is a recipe for disengagement. Falling short of customer expectations is like being savaged by a Demogorgon: your brand may never come back. Chances are, your viewers won’t return either.
User experience is everything. But to achieve it, you must start in the right place. Many service providers mistakenly believe that the Content Delivery Network (CDN) is the engine room of user experience, and so they narrowly focus their attentions there. It’s an upside-down approach that poses significant risks to the brand experience. There’s more to delivery than designing an app and configuring the CDN. The drivers of user experience are determined far earlier in a complex process that bridges multiple players, locations and viewer-specific variables. It’s a monster. But it can be tamed.
Taming the monster
There are four core areas that must all be optimised to deliver video, at scale, directly to the consumer.
#1: Platform architecture
OTT infrastructure is multi-faceted and involves multiple vendors, skillsets and technologies that all add value across the delivery chain. With so many actors in the process, it’s important to map everything out up front. Often, however, service providers don’t ‘own’ the process but instead take best practice from all their vendors and hope everything comes together in the end. Unfortunately, it rarely does. Poorly conceived, piecemeal planning generally leads to suboptimal architecture and inefficiencies that reveal themselves in the user experience. These inefficiencies are magnified when providers try to scale delivery to mass audiences across multiple countries.
A good plan will set out the vision and map the vendors and technologies required to establish the optimal architecture. Best led by a technical architect, it will outline how all the components inter-relate and provide a foundation for the processes needed to (quite literally) deliver the vision. The roadmap doesn’t need to be set in stone, but it does need to be in place from the beginning. The best plans are proactive, adaptive and iterative – but they’re all anchored to a defined vision.
#2: Media preparation
This is central to the user experience. Building on the roadmap, media preparation is about understanding and optimising the various technical components required to deliver high-quality video at scale. In a complex world with no single encoding standard, preparing content for delivery requires a granular understanding of a range of variables; devices, screen sizes, bitrates, frame rates, sequencing and configuration. It’s highly technical and often goes beyond the skillsets of a technical architect. It requires the input of a true video expert. The slightest misalignment will inevitably lead to poor quality streams and suboptimal user experience.
#3: Apps and Control Planes
The foundation for building effective applications is, once again, enshrined in the roadmap. Unfortunately, applications are often designed by great software developers who are, sadly, unfamiliar with the peculiarities of video delivery and the idiosyncrasies in how users browse and view content. Simple things like the order in which components load in an app can massively impact the speed of delivery. The development of intelligent rules around aspects like pre-buffering or pre-caching content can help create a smoother experience. Conversely, when apps aren’t optimised for the way your services work, it can kill the experience altogether.
Control Planes also dictate the quality of the user experience. The Control Plane sits in the background, discretely validating user entitlements, authorisation, location and licensing rights to make the app work quickly and effectively. However, a poorly configured Control Plane can be the Shadow Monster of video delivery – significantly delaying or interrupting the user experience. For example, if 100,000 people hit the same server at the same time – and, each time, the system must go through a convoluted series of steps before serving content – the experience will lag and the viewer will become frustrated. That spinning wheel on the player at the time of playback isn’t only as a result of CDN interaction. Once again, configuring the platform to avoid delay requires specific knowhow and expertise.
#4: Media delivery and playback
Our golden egg – this represents the delivery of the product to the consumer – but it’s still so easy to get wrong. It’s frequently assumed that the CDN is the only actor in the delivery phase and that delays in playback are entirely the CDN’s fault. However, as outlined above, there are numerous steps in between that can cause delay and ruin the experience. It’s important to remember that a CDN can only start playing video once it’s had a request to do so. If that request takes too long, it’s commonly down to misaligned processes or suboptimal sequencing earlier in the delivery chain.
The CDN is the endgame of video delivery and a crucial determinant of user experience. However, whilst its configuration undoubtedly needs optimising, it shouldn’t be your first area of focus. Taking an upside-down approach to designing OTT infrastructure will invariably mean you overlook common drivers of inefficiency – triggering user experiences that are spoilt by stranger things. To avoid them, it pays to partner with experts who know how to build successful video services in the Cloud and can help you journey to a parallel dimension. What’s more, unlike Jim Hopper, they’ll do all the groundwork and won’t leave you trapped in the reeds.