Pivoting To Video: Lessons Learned
Monday, October 14, 2019
Posted by: Laura Dawson
In February 2016, after a career spent wholly in the book industry, I made my own pivot to video. Through my publishing contacts, I landed a consulting job at HBO in the Metadata Management and Taxonomy department, which turned into a full-time position.
The leap was not as drastic as it might seem. Metadata, after all, is pretty much the same wherever you go – a catalog contains titles, descriptions (both long and short), contributor names, genres and categories. My skills were quite transferable.
But of course, broadcast and cable media is a different industry with a different business model – and that translates into some key differences. What I experienced in my time at HBO were some of the most interesting problems I’d ever tackled, metadata-wise.
Just as the book industry experienced a massive upheaval with ebooks, the entertainment industry experienced a similar disruption with the advent of streaming. HBO’s foray into that technology began in 2014, with the introduction of HBONow on Apple TV.
The impact of that project resonated into its databases, which had been built to support scheduled programming transmitted by satellite to cable affiliates. The change in business model (while continuing to supporti the traditional cable model) meant the creation of new systems that had to interoperate with the old ones.
That change resulted in a lot of mapping, new tables in old databases, and new ways of thinking about the content being aired. Longstanding distinctions between products and promotions, for example, blurred into “program assets”, encompassing any piece of video content. Evolution in technology – packaging audio, video, and text components into a package rather than hard-coding them in a single file – meant that the way the company thought about distributing that content had to change.
The book industry’s embrace of ISBNs and DOIs (and potentially ISCCs) leads me to think that situations where assets are blended (text, video, audio) would be more easily navigable. We see these instances especially in the educational sector, where textbooks often have interactive media components. This has been a real problem in the video industry; when content is chunk-able, they don’t have reliable identifiers to disambiguate those “chunks”. The book industry has been talking and thinking about chunking since at least 2008.
A key difference between the book and cable industries: both historically and today, the video industry is deeply invested in tech. HBO, in particular, spends quite a bit on tools like photogrammetry, cameras, drones, and special effects. Game of Thrones pushed the company into experiments that furthered the industry as a whole, and upped the game of future HBO productions with a wealth of knowledge and experience derived from those experiments.
The book industry does not feel such an intense need to continuously leapfrog over past technological implementations – the codex just doesn’t change that much, whether digital or print. Once publishing nailed standardized data feeds and ebooks, technical innovation plateaued, and much of the work entered a maintenance stage. As a result, book industry metadata became relatively stabilized.
Video metadata, however, has to support the constant, rapid technological innovation. It can feel overwhelming at times, but it certainly forced me to learn quickly. It’s possible that new formats and channels might push book publishers to engage in more frequent technological innovation. The growth of voice-assisted technology and audiobook sales could signal such a shift.
Another key difference between industries is the lack of metadata centralization in video. Unlike most book industry organizations, HBO does not have a single source of truth – a centralized database. Every time it attempts to create one, rapid innovation outpaces it and another system has to be set up. Because of the pace of the business, it’s hard to take the necessary time to synchronize the flow of data to create a definitive repository. I spent most of my time creating mappings between databases, and creating standard vocabularies that could be pulled into systems on the fly.
A factor in the centralization issue is the lack of a universally adopted identifier. The book industry was far, far ahead of its time in standardizing around the ISBN. Having that identifier as the backbone for databases results in more organized, cleaner metadata overall. This is the flip side of more frequent innovation: if it comes to book publishing, it’s critical to keep things like universal identifiers, or managing metadata will grow more expensive and less reliable.
While some video companies have adopted ISAN (International Standard Audiovisual Number) and others have adopted EIDR (Entertainment ID Registry DOIs), there’s no industry-wide agreement on this issue. It shows up most directly in the lack of system uniformity. Believe it or not, that’s one area where the book business has the upper hand over other entertainment and education industries.
Finally, there’s the level of investment. In the video business, money gets spent on projects that never see the light of day, at all levels – the business has that luxury. For every show that makes it to air, there are dozens stuck in “development hell” – optioned, but never filmed. This is true even though the cost of developing a video asset almost always dwarfs the cost of bringing a new book to market.
Post-broadcast methods of recouping that money are vast and varied – traveling costume exhibitions for Game of Thrones or Downton Abbey, licensed Funko-Pop dolls, and ComiCon souvenirs are all examples. For years, HBO even had its own retail shop, in the base of its headquarters. For product sales and give-aways, all of that metadata needs to be catalogued and tracked as well. The Archives department does a booming business in preserving information about bygone props and materials, for example.
In all, my experience moving from the book industry to video entertainment has been enlightening in a lot of ways. I have felt rather like Cousin Greg from HBO’s show Succession: goggle-eyed at the goings-on. It’s been a real education.
This article is brought to you through a partnership with Amnet, a technology-led provider of services and solutions, catering to the needs of businesses for content transformation, design, and accessibility. The points of view expressed are those of the author and do not necessarily represent the perspectives of Amnet or of BISG.
Laura Dawson spent 30 years in the book industry, focusing on metadata, before moving to the media/entertainment business. She currently works at EDO, Inc., a data analytics firm that measures consumer engagement and intent in media advertising.