For traditional television broadcasters, revenue opportunities are shrinking. Earlier in 2017, subscriber growth fell dramatically. The industry lost a record 762,000 subscription customers in the first quarter of 2017, according to PwC. What’s more, ad revenue was expected to decline by more than 5% by end of 2017. TV operators are being affected by growing competition with digital on-demand providers, as well as content owners using OTT services to distribute media themselves. In this piece for ExchangeWire, Srinivasan KA (pictured below), co-founder, Amagi, explains why TV operators should move away from satellite and fibre-based content delivery models and embrace AI. 

To optimise revenue, broadcasters need to be inventive. In fact, I believe the industry should consider embracing an approach currently being used with great success in the manufacturing sector. For years, durable goods manufacturers have seen a decline in orders brought on by global competition. To counteract the dip, manufacturers have cut costs and protected margins by overhauling their supply chain infrastructure. Many are adopting new technologies to automate and streamline traditional operations, driving billions in savings in both capital and operating expenses.

Broadcasters should take a page out of this book. From content management, to playout, to delivery, they need to rethink the onerous and expensive traditional TV infrastructure and media supply chain. Doing so can substantially raise revenue and help in rapid expansion of business.

Here’s how.

Stop relying on satellites alone

The primary reasons to rely on satellite or fibre have been the reach and reliability of these legacy delivery and hardware-based systems. Unfortunately, this model is no longer sustainable, as TV networks launch more international channels to attract audience. The capital-intensive nature of these methods, and manual workflows, simply do not fit in with the lean model that TV networks need to make channel delivery economically viable. Enormous hardware is required to run satellite and fibre technology. Physical data centres and on-site personnel are also needed wherever content is transmitted. All of these expenses cut deeply into broadcaster revenue, particularly for mid-size operators. As subscription and ad revenue shrink, these costs become harder to navigate.

Srinivasan KA, Co-Founder, Amagi

To overcome this, savvy broadcasters are embracing new technologies. VICE, for example, is now incorporating cloud technology into channel content delivery, instead of relying solely on traditional broadcast models for its international expansion. After several years of innovation, the cloud has emerged as a truly viable alternative for satellite and fibre-based TV content delivery. Cloud technology not only makes it viable to air linear TV channels, but also adds value by improving transparency of the entire workflow. With cloud technology, end-to-end broadcast operations can now be controlled and monitored remotely – without a physical hardware presence or personnel for operations. The ability to scale up on reach, while scaling down on costs and resources, is a massive gain for TV networks. I can say with certainty that the role of cloud technology as an agent of transformation can no longer be ignored by the broadcast world.

Start embracing artificial intelligence

While cloud solves the delivery part of the content puzzle, TV networks today are also grappling with how to process large volumes of content. Somehow, the manual nature of content processing has never been questioned – even today. As TV networks aim to broaden the range and diversity of their content, they need to process massive volumes of data before they can use it for playout. This is where we must seriously consider applying artificial intelligence.

For example, manually creating segments from raw video data is time-consuming and tedious. It’s also prone to error, which can have a huge impact on the final production, as well as the reception among audiences. In the traditional system, a manual specialist is engaged full-time to create and tag segments. Protecting the integrity of media assets and ensuring that the right asset is played at the right time – and with the right parameters in place – is a major task. Now, imagine this process across different channels and geographies! With recent advancements in machine learning, the same process could be done much faster, and with fewer resources.

Now, AI and machine learning tools can intelligently identify snackable segments from a longer piece of live or pre-programmed content. The technology has advanced to the point where it can create tags to store these assets efficiently. It can also use pattern recognition to perform advanced quality control, detecting abnormalities like blank frames or freeze frames, faulty audio, and more. Most importantly, though, AI can be used to generate these smaller segments for deployment beyond linear TV. AI can analyse video, and identify and create segments that would be of interest to online viewers. A highlights package, for instance, could be created for a football game via machine learning versus human intervention. AI can even pick out and use the ‘in-play’ moments of a game, navigating commercial breaks, and delays in the action. Similarly, a programme could be converted into webisodes or promo clips. AI’s potential to help broadcasters make more money is boundless.

Broadcasters are experiencing tremendous change and fierce competition. But technology can help them navigate the changing environment. Moving away from a satellite and fibre-based content delivery model is one step. AI is another. By embracing these new opportunities, the industry can dramatically improve ROI, and deliver greater success as it moves into the future.