Interoperability is a big challenge in streaming workflows. A lot of resources are spent creating connections between different elements that are natively integrated. For example, developers may be tasked with building custom middleware to connect encoders with operational monitoring systems so engineers can know the status of encodes. This needs to happen dozens of times within an OTT architecture. That’s because for streaming operators to provide the best possible service, all of the independent components must be integrated in some fashion.
“Without native integration, it would be impossible to create a workflow through which content can be acquired, encoded, secured, and delivered efficiently.“
Even though APIs can make this integration easier, it still requires extra work. Unlike traditional broadcasting, there are no government-mandated standards on how components should integrate together. There is no such thing as a “streaming system” like there is a “broadcast system.” As such, each OTT architecture is often different with components connected in different ways, and operators must dedicate resources to managing and building those integrations.
Interoperability platforms simplify OTT architecture management
Over the past decade, we’ve seen the growth of “interoperability platforms” such as Zapier and Segments. These platforms enable developers to create integrations between different systems using vetted and managed API connections. They also allow them to connect third-party systems together easily and seamlessly, which is a major advantage compared to the Do-It-Yourself (DIY) integration approach.
Additionally, these interoperability platforms save countless hours of not only building integrations but also managing and updating them. In contrast, organisations that develop the integration between streaming workflow components themselves face a huge resource drain. They must manage vendor APIs and account for their own systems as they change, both of which necessitate more time and cost-intensive custom coding.
When using a Segment, Zapier, or other platform, OTT operators no longer need to develop custom middleware as the integration is already handled. This not only significantly reduces development time, but also greatly enhances stability and reliability. The platform operator manages the interoperability, ensuring connections remain up-to-date and operational. With such an ecosystem of native integrations – which are developed by the system owner and maintained by the interoperability platform operator – companies can focus more on other aspects of their business and less on integration. All the while, they’ll know that the component pieces of their system are scalable and resilient.
Leveraging native integrations for analytics components within the OTT architecture
Streaming, of course, isn’t like many other systems. There just isn’t an interoperability platform for streaming architectures. However, while the entire OTT architecture can’t be driven by platforms with native integrations, there are functional areas, like analytics, where streaming operators can make use of systems in which integrations are managed by someone else.
When it comes to streaming video analytics, operators have become more sophisticated in how they manage the flow of data through their architectures. Where they were once only collecting data from an endpoint, they are now collecting it from a wide variety of systems throughout the workflow. Some of these systems, like the player, are owned by the operator, while others, such as commercial CDNs, are not. This makes connecting to them, gathering data from them, and integrating them into existing operational dashboards far more challenging.
“The truth is that - whether it’s acquiring data, sorting data, or visualising it - every OTT operator struggles with analytics.”
In that sense, the main challenge is how to best make use of the data to improve the streaming workflow and ensure the best possible quality of experience. However, to do so, the right data must be available to start with. To gather all the data they need, from the systems throughout the workflow, operators often have to build custom integrations between workflow components and their analytics platforms. And that is exactly the core issue: building and maintaining integrations isn’t their business, operating a streaming service is. What operators truly need is an analytics system in which the platform owner provides native integrations so that connecting one workflow component or a dozen requires the same level of effort. The work of integration is already done and the OTT operator simply needs to enable the connections they want.
So what makes an analytics platform with native integrations so important? Below are the three main advantages.
Data normalisation
One of the issues with creating your own integrations with third-party providers, or using the provider’s integration, is that you aren’t in control of how the data is ingested into your analytics systems. Each integration must be addressed separately and the data normalised individually.
In contrast, when the analytics component in a provider’s workflow is built on an interoperability platform, that normalisation is handled by the platform, thus ensuring that every new data source added through the platform conforms to the OTT provider’s requirements. This reduces data post-processing time by conforming data points to the operator’s data dictionary before it hits the visualisation tool. Less post-processing time means being able to more quickly use the data for its intended purpose (i.e. root-cause analysis).
Operational standardisation
Without assured interoperability, adding a new source of data often means a new visualisation tool. That requires operational personnel to learn how to read the new data and, more importantly, connect it with other data sources. This radically increases the time it takes for root-cause analysis and other QoS/QoE interventions.
However, when data sources are added natively via an interoperability platform, data can be pulled into an existing tool that operational personnel are already using. That means no more re-learning and a standardised way to approach data analysis.
Improved data mining
Combining different data sources helps reveal both powerful insights into performance and QoE as well as monetisation opportunities, such as behavioural and viewing patterns. However, without an easy and standardised way to add new data sources, connecting them is very challenging.
By utilising an interoperability platform for the OTT provider’s analytics component though, data mining is significantly improved. Operators simply need to select an available integration, specify configuration settings, and let the platform do the rest to see new connections within the operational visualisation tool.
Native integrations allow you to focus on QoE instead of worrying about your OTT architecture
For a streaming platform operator, content and experience are the main differentiators, not the underlying OTT architecture. Time spent integrating or managing third-party technologies and services is time not spent on ensuring your viewers have a great QoE. This issue is amplified as APIs and systems inevitably change over time, which would require yet more resources spent on maintenance. In the end, there are only so many development and operational resources to go around. It’s not tenable to task them with building and maintaining your video streaming technology stack, integrations as well as quality of service.
Of course, analytics is just one example of how an interoperability platform, with an ecosystem of native integrations, provides significant benefits. There are others though. In fact, many OTT providers are already using some of them, like AWS, which has native integrations to Elemental, Cloudfront, S3, and more. Although not exactly the same, as all those services are owned by Amazon, it’s still a step in the right direction. Interoperability platforms, even such as those for analytics, offer a way to mitigate the inherent fragmentation in systems like streaming video architectures, systems composed of multiple components from a variety of sources.
Maybe one day there will be a single platform for the entire streaming workflow. A “streaming system” through which all integrations are managed by a third party, freeing the OTT operator to focus solely on the viewer experience. Until then though, streaming operators can leverage native integrations to conserve valuable technical resources for what really matters: the viewers.
