How Good Metadata Can Make You Money
With so much content competing for viewers’ attention, it’s no surprise that one of the top challenges for media owners and distributors is how to ensure consumers find and choose their offerings over those of their competitors. Viewers are drowning in choices and—thanks to Google—millennial audiences, in particular, are conditioned to expect instant gratification, yet their ability to find engaging video content frequently misses the mark.
An organization may have great content that critics adore, but the ability to monetize a back catalog is directly related to viewers’ ability to find what they are searching for. And when these fickle viewers aren’t sure what they are in the mood for, the ability for a platform to offer compelling suggestions based on genre, favorite actors, locations, ratings, and reviews can make the difference between abandoning a platform and the brand temporarily and seeking satisfaction elsewhere.
Metadata to the Rescue
Metadata is the key element that successfully leads viewers to your content. The richer the metadata, the more effective it is when uniting viewers with what they are looking for. A title without an engaging synopsis is begging to be passed over. Yet, so much of today’s available content lacks even the most fundamental descriptions.
By focusing on consolidating, organizing, and enhancing the quality of their metadata, content owners can optimize the value of every asset in their media library. OPEX costs are reduced and flexibility is enhanced. Good metadata goes beyond “discovery” for viewers; it also enables workflow automation and smarter operational insights that can inform organizations on the strengths of their catalog, and where to best make future investments.
Metadata Doesn’t Grow on Trees
Where is all the metadata for my content? If it’s content I have the rights to distribute, why don’t I have it already?
Different types of metadata exist, in various forms, along the content creation and distribution pipeline. From the moment a script is commissioned, metadata is being collected. Once a show is in production, mountains of metadata come in from the file-based cameras including shoot date, location, take number, scene, reel, etc. Depending on the production, loggers are brought in to tag and further categorize the hours of footage being captured. All this metadata is critical to those in post-production who will organize and edit the show. However, much of it will be discarded as the show progresses from system to system, through multiple processes, on its journey to becoming a final show master with a multitude of deliverable versions.
An entirely different set of metadata is required to properly identify and categorize media files that are ready to be distributed. There is descriptive metadata—which includes the title, synopsis, release year, cast and crew, thumbnail images, etc.—and technical metadata, which includes availability windows, buffer windows, program segmentation information, video profile info, and much more. If the distribution is worldwide, then the descriptive metadata needs to be available in multiple languages as well.
Where Does All this Metadata Come From?
Large established companies frequently have descriptive metadata held in different databases from their technical metadata for the same media. Linear metadata from archived programs that have been converted to file-based assets may live in yet another database. It’s common for metadata to be sitting on multiple media asset management (MAM) systems involved in the creation of these programs, with no clear way to combine, normalize, and de-duplicate it. This is one reason why metadata increasingly comes from metadata vendors who specialize in collecting, collating, and transforming this information for the very companies who originated it. Companies like Gracenote, TiVo (recently Rovi), Ericson’s Red Bee, IMDB, and Rotten Tomatoes, all sell metadata as a way to enhance the entertainment discovery experience. In some cases, these metadata vendors actually sell back some of the same metadata they’ve bought or acquired from the original content owners.
What the World Needs Now is Metadata, Sweet Metadata
What’s needed, now more than ever, is a way to give broadcasters and service providers the ability to consolidate their existing metadata into a single dataset, while having the option to aggregate further content specific metadata from third parties. Most importantly, content owners need to be able to do this in the most automated way possible to further streamline the distribution of metadata to distribution partners, wherever they may be, and in whatever language they require.
High-volume “metadata injection” requires state-of-the-art machine learning techniques to build associations and match metadata from disparate sources. The machines need to score their confidence in matches they’ve made, flagging human operators when confidence is below a set threshold.
There’s an entire metadata ecosystem developing and transactions between metadata vendors need to be centralized and simplified. Applications need access to APIs, so they can programmatically grab information without human intervention. Business rules and localized conventions need to be respected and enforced, minimizing mistakes in syntax and conventions across geographies. Any system capable of aggregating metadata must also be able to transform the data correctly for its intended destination. This gives organizations the ability to manage the complete metadata lifecycle, leading to a more organized back office, improved ability to match and aggregate content, and an enhanced quality of experience.
[This is a vendor-contributed article from Piksel. Streaming Media accepts articles from vendors based solely on their value to our readers.]
Metadata can do far more than help publishers manage assets and customers find content. Employed the right way, it can help deliver a personalized viewing experience that improves ROI.