A new wave of disruption is taking place in the world of TV metadata and it’s called AI. Will this collection of technologies be able to fundamentally transform the TV business and solve the operational workflows that revolve around managing TV content archives in the future? And can it help us solve the industry’s huge content discovery challenges? Read the full article, “Metadata is the Answer - and the Problem” in the Fall 2019 edition of OTT Executive Magazine by Janet Greco.
The Backstory
It’s useful to put in context that once we lived in a simpler age with just a few channels. Content discovery wasn’t a thing. There were no major issues. You could access your TV listings in a small section of your newspaper or get all the detail and feature stories from dedicated TV listings magazines.
It’s nearly 30 years since the first wave of innovation happened in the distribution of TV listings when we went from paper, to electronic. Since then, not too much has changed in the manner of data exchange between aggregators and operators.
When the number of TV channels surged in the 80s as a result of cable and satellite delivery, together with the deregulation of TV in Europe, grid formats were introduced in printed formats, and scrolling guides would let you, in the space of a few minutes get a sense of what was on right now.
As with the grids, Electronic Program Guides introduced in the mid-90s made it a bit harder to access program information, because you had to click into the listing to view the description. Nevertheless, you could more or less find your way around. The advent of EPGs was a natural evolution of the explosion of channels on pay TV platforms around the world.
Driving these guides was the TV metadata consisting of time, title and a program description. In those simpler times program listings were not transmitted electronically. In the early days it was phone calls, telexes, faxes and good old postal mail.
Corresponding to the growth in channels and their corresponding international distribution, the old ways quickly became unmanageable.
The first disruption TV listings distribution took place in the early 90s with the advent of “electronic delivery” services. These intermediaries, which came to be known as TV metadata aggregators, provided a modern way for channels to deliver their listings to the wide number of publications, later websites and EPGs in an efficient manner.
TV aggregators took over the burden of normalizing the many different file formats received from the channels. It could be Word or Excel files, Pagemaker Files – everything under the sun – with program updates coming by fax. It was up to the aggregator to “normalize” these files into a useful format. After all, if a cable operator needed the listings of 100 channels, they’d require a single data import format, not 100. Metadata aggregation therefore acquired an essential role.
In a nutshell, pretty much, things have remained the same. This data is still acquired in multiple formats, aggregated, normalized and then be delivered to anyone who needs to provide a guide product, eg, the EPGs and content discovery systems.
This work flow remains to this day dominated by behemoth TV metadata aggregators. These businesses grew large because of the commoditization of the data and the acquisition of a numerous smaller players over the years. The hope for integrated systems within legacy broadcast and pay TV operators has remained just that.
Data Silos
One of the key concerns for pay TV operators and broadcasters is how they are going to maintain their competitive advantage over the constantly proliferating number of OTT and D2C services.
Every operator wants to deliver correct and consistent metadata and be able to surface that information to their audiences, giving them a good user experience and the ability to discover content that will delight their audiences.
But there’s a problem. Metadata management issues persist across many legacy TV enterprises, and the resulting impact is felt across the entire digital ecosystem, ultimately impacting the consumer experience negatively, not to mention the bottom line.
Across pay-TV, broadcast and OTT, data management takes place in fragmented silos built up from legacy systems used to manage different aspects of TV operations. It is not so easy to knit all these sources together. These TV operations, ultimately have to acquire the metadata from somewhere, or normalize it themselves if it is their own.
For many years the persistence of a departmental “we’ll do it our way” mentality gave birth to a mega-mess of multiple IDs, multiple data sets and siloed work practices based on legacy systems with no easy way to de-duplicate and harmonise the data.
These legacy systems are no joke to sort out. We have been talking about it for approximately forever. But we are quite stuck now. We expect dashboards, analytics and integrated data whose metrics can be interpreted well in order to make better business decisions.
This problem pertains mainly to the digital non-native companies and so the multiple siloed system aspects are largely absent when it comes to the incumbent OTT services, but they also have to get their metadata from multiple operators and deal with multiple formats that have to be normalized (still).
To what extent can the introduction of AI-driven tech help solve the operational challenges for legacy players? Read the full article to find out. You can also learn more at our upcoming IET Media event: AI and Metadata: Engaging Audiences with better Content Discovery and Recommendation
Get your free subscription to OTT Executive Magazine here. Read the full current edition here.