Scaling OTT architectures for a multi-device world
The new connected device landscape provides an exciting opportunity for content distributors.
Over-the-top (OTT) video is now mainstream. The challenge lurking behind this phenomenon is that consumption is moving toward a highly fragmented set of devices, and no single solution exists to solve the operational and technical complexities of delivering video to a diverse landscape of devices.
When OTT streaming services launched in the mid-2000s, video delivery was simple and straightforward: single-bit-rate, Flash-encoded content streamed to desktop browsers.
Fast-forward to today, and OTT services have expanded to smartphones, tablets, gaming consoles and Blu-ray players. Within the next two years, IBB Consulting believes that the number of these devices will grow 50 percent, reaching a point where every person in a household will have access to a connected device.
Device deviations range from the obvious, such as the user interface, remote control and screen size, to subtle technical deviations, such as supported codecs, streaming protocols, programming language and embedded DRM.
The range of device deviations, further complicated by stricter authentication requirements, has significantly changed the solution architecture. Multichannel video program distributors (MVPDs) face both engineering and operational challenges as they try to keep up with the latest standards, formats, platforms and form factors.
However, no single solution exists. Consequently, content distributors need to make decisions around which devices to support, what kinds of services to provide and how to best position a service against a competitor’s offering.
Additionally, content distributors should anticipate that this fragmented market will continue to persist in the coming years. They should therefore follow some important design principles when developing the appropriate delivery infrastructure and operational processes in order to support a fragmented device landscape.
Three key OTT challenges
Content distributors that seek an “any device” solution face three key scaling challenges: building and maintaining custom applications for each platform, managing the upstream content ingest, and enforcing the streaming rights for each piece of content in their catalog.
Building and maintaining custom applications
Initially, HTML5 held promise as a one-size-fits-all option. However, its current inability to provide adequate content security renders it unsuitable for streaming high-value content. Consequently, distributors cannot use the traditional browser-based approach often employed to reach desktops. Instead, they must create native applications to reach the connected devices on their roadmap.
However, creating and maintaining a client application for each supported device is a process that scales inefficiently. Despite sharing many product features and visual design elements, each new client application and subsequent release must progress through the application lifecycle independently (see Figure 1).
Developing separate applications across a wide array of devices requires many different technologies, including application languages, player frameworks, security technologies and software development kits. This means that very little application code can be reused across platforms. It also requires a diverse set of development skills, which increases the challenges of maintaining these applications. Furthermore, diversity of form factors and client features means optimizing the user experience per device, requiring further variation among client applications.
In addition, when content distributors create a new client application, or update an existing one, they need to review the proposed security framework with each of their high-value content owners, as well as go through the respective app store certification process. Together, these security audits and certification steps are time-consuming processes that slow speed to market and introduce additional costs.
Managing the upstream content ingest impact
Variability in downstream technologies on the connected device has significant upstream ramifications on ingest workflow (see Figure 2). The choice of video bit rates, streaming platform and DRM determines the way source video will be transcoded, segmented and encrypted. Each new transcode profile, streaming protocol or DRM solution employed in the client application creates a new branch in the ingest workflow. This can have a multiplicative effect on the total number of files created and published to the content delivery network (CDN).
The end result is a more complex ingest workflow, greater hardware demands and increased storage costs.
Rights management and user authentication
As MVPDs have expanded offerings to connected devices, a fragmented set of rights agreements and underlying distribution requirements have emerged. Variability in agreements can result in differences in viewing windows, device registration, geo-restrictions and concurrent streams. To manage this, content distributors will now need to track and enforce diverse entitlement rules on an asset-by-asset basis.
Five engineering principles
Delivering to a variety of connected devices requires the development of a scalable technical architecture that supports today’s fragmented and rapidly evolving landscape. To do this, content distributors should focus on five key guiding principles: limit client-side application logic, design flexible ingest solutions, develop a solid content rights management system, consolidate delivery technologies, and evaluate and adopt emerging standards.
Limit client-side application logic
By limiting client-side application logic and maximizing functionality in the cloud, content distributors can create cross-platform efficiencies. To do this, content distributors should push business logic – such as video entitlement, stream concurrency and device management – into the cloud.
Additionally, content catalogs and linear guides should be rendered from the cloud instead of being downloaded directly to the client. All of this should be managed through a series of defined APIs between the client application and the functionality in the cloud.
Finally, administrative functions like registration, device management and account maintenance can be built exclusively for the desktop site instead of being incorporated into each client application.
Design flexible, automated ingest solutions
Flexibility and automation should also serve as key design axioms for ingest solutions. The ability to quickly create new ingest workflows that can systematically convert a mezzanine video asset into new delivery formats will save both time and money over the long term. For example, some distributors are already moving toward “on-the-fly” solutions that postpone transcoding, segmenting and encryption until a user requests playback of the asset.
This approach provides maximum flexibility to add or remove ingest technologies while reducing storage costs and enabling stronger security in the form of session-based encryption.
Implement a robust CRMS
Content distributors must implement a content rights management system (CRMS) that is capable of managing the disparate terms and conditions for each video asset, as well as scaling to support real-time awareness of each user’s activities and entitlements.
Client applications will need to provide information about the user’s state, including geolocation, device ID and access network. The CRMS must then make complex entitlement decisions based on the user’s entitlements and the content restrictions.
In addition, to create a comprehensive CRMS, content distributors should also pursue broad rights within content agreements that standardize entitlement decisions across platforms and content partners. Standardization across rights agreements will simplify both the solution architecture and the operational support processes.
Consolidate delivery technologies
Consolidating delivery technologies simplifies workflows, reduces hardware requirements, and limits origin and archive storage. This means that content distributors should reduce the number of transcode profiles, streaming platforms and DRM solutions employed in client applications.
Content distributors must carefully consider each new transcode profile they introduce into their ecosystem. Choosing resolutions that can be shared across platforms and devices can significantly simplify ingest workflows with limited impact on the user experience. Similarly, aligning on one or two streaming platforms can provide significant cost savings and operational efficiencies.
Lastly, the choice of DRM is another important decision in designing a scalable solution. Choosing a proven DRM with widespread adoption has many advantages. One key benefit is allowing distributors to reach many devices with a common set of files. This translates to upstream simplification of ingest workflows. It also reduces custom development to integrate DRM clients with native or third-party players and accelerates the security approval process with content owners.
Evaluate and adopt emerging standards
The OTT landscape is a complex space with many competing proprietary standards. Instead of developing custom solutions, content distributors should leverage emerging open standards such as HTML5, MPEG-DASH, DLNA and ADI 3.0, which are designed to improve the scalability of a cross-platform, multi-device video service.
Building an architecture that is prepared to leverage these standards as they are adopted by consumer electronics manufacturers will simplify some of the future challenges of a TV Everywhere solution.
The bottom line
Content distributors should clearly understand the cost of creating and maintaining client applications for each new connected device on their roadmap. Content distributors must also weigh the types of experiences that they wish to offer and appropriately balance this with their content distribution deals. Armed with this data, content distributors can more effectively weigh their costs against the strategic and business benefits of expanding to new connected devices.
The new connected device landscape provides an exciting opportunity for content distributors to reach end users on any device, anywhere, anytime. However, achieving this vision requires content distributors to acknowledge the fragmented nature of the connected device market. By understanding the nature of these challenges and applying a few key technology principles and sound business analysis, content distributors will be better prepared to realize their TV Everywhere vision in a cost-effective and operationally efficient manner.
Emails: firstname.lastname@example.org | email@example.com