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.