As the Video-on-Demand (VOD) market becomes increasingly competitive, the customisation of the viewer experience has become a key differentiator for media organisations. Viewers have come to expect a highly personalised service: the importance of both great User Experience (UX) and a simple path for content discovery can’t be overstated. Linear channels are also storing more content assets, both archive and new episodes, in the cloud. To ensure content supply chains are running efficiently and to streamline scheduling, content owners need to efficiently tag and track key information about their media.

Metadata plays an important role in this, but there is often a feeling of "more is more", a sentiment that can and does cause challenges further down the line. The over-proliferation of metadata collection can obscure the objectives for gathering it in the first place, but the process of collating ‘good’ metadata is not always simple. There are a number of challenges that arise when collecting and managing metadata on a mass scale, such as informational silos, the needs of different content supply chains, and metadata standards. If not managed correctly, these issues can all feed into a sense of metadata fatigue.

The role of metadata

Put simply, the purpose of metadata is to enable media businesses to make the most out of their content, whether that content is used for on-demand platforms or for linear broadcasting. Content owners and broadcasters now have unimaginably huge catalogues of media that need to be identified and categorised for effective search and discoverability. In addition, now that older content from the archives is being leveraged for monetisation purposes, there is a huge need to identify and categorise not only new media assets, but also older material in the archive. This process will benefit both the end-consumer and the media operations user.

Metadata also has an important role to play in linear broadcasting, when broadcasters need to change scheduling at the last minute, or to remove certain content, or pull whole shows. This may happen when there are sensitivities over a particular political incident or recent news event. As well as being crucial in those kinds of time-sensitive situations, effective metadata is also vital when compiling a collection of particular content from an archive. This could be a compilation focused on a specific person or a theme. Let’s say an actor wins an award and a broadcaster wants to collate past appearances, interviews and other notable points relating to the actor’s life. To do that quickly, metadata needs to be accurate, effective and searchable.

Metadata management

Media businesses typically have lots of different types of assets, each with a range of associated metadata, and managing this is, unsurprisingly, a complex process. In order to manage a media-rich archive at a granular level, media companies really need a MAM system in place to index their catalogue successfully. Then they can run searches, screen the archive, and do things like marking clips and captioning audio descriptions for tagging. Without effective management of the search, annotation and indexing processes, the media archive will quickly become unstable and siloed.

If the metadata recording system isn’t integrated with the whole media production process, there might be a need to use different systems to track different types of content, which is obviously unsustainable because the archives are all interrelated. Not only is this inefficient in terms of duplication, but it can also lead to the proliferation or duplication of metadata about the same person, object, or theme.

Without using a MAM with an intuitive search capability that enables the user to handle, mark, track, and publish rich metadata at scale, the metadata is going to be an inconsistent mess. In that scenario, you would likely see some metadata hosted in one place, some in another place, some that's only attached to the subtitles or audio description files, some that's only attached to one episode in a series, and some attached to each and every episode. It is critical to have a strong search capability at the core of any MAM system.

Good metadata vs. bad metadata

Good metadata is useful metadata, which has a clear purpose and has been collected with clear objectives in mind. Good metadata is about getting to the core of what is really needed. Some metadata is needed for the user to easily discover content, while other metadata is needed because it is useful in the MAM for the media production process.

It is fair to say that any metadata is useless unless it has a purpose. If unnecessary metadata is being logged, it can obscure the useful metadata, making the search process more difficult. In this way, unhelpful, excessive, or duplicated metadata can actually become a hindrance.

Informational silos

There is also an issue of metadata not always being shared both interdepartmentally and also, perhaps more understandably, from one media company to the next. The act of creating metadata is in itself an investment, so there is a reluctance to share metadata with other media companies unless it can be monetised. As content passes through the media chain, the generation of metadata can sometimes be duplicated by different media companies.

Even within a single media company, it is not unusual for some teams to use one metadata recording system then not share that system or metadata with other teams, so the work is duplicated. These high levels of duplication are not efficient for individual media businesses, nor for the industry as a whole, and the terminology used can vary from team to team.

Different content workflows have different metadata needs

If metadata is being recorded for lots of different functions such as indexing, archiving, compliance and QC, it can very quickly proliferate, even if it is being recorded with clear purpose. To get round this, metadata processes need to be made more efficient and more streamlined.

Processes must set out an efficient way to mark, track and publish technical metadata, descriptive metadata, custom metadata, customer specific metadata, and supply chain metadata. Carrying metadata from the point of ingest, to the end product on the MAM platform, is probably the most efficient point in the media supply chain to record metadata but what’s recorded needs to work for all media operators involved in content processing.

Empowering media operators to work efficiently

It is important to empower media operators to use their skills, work efficiently and give the best insights to teams at later stages in the media supply chain. Ideally, a metadata system or process shouldn’t allow operators to take actions that are unhelpful to the metadata process, but this is not always the case. What is important is cooperation, collaboration, and communication. AI tools can be used to assist media operators in tagging of descriptive metadata, to free up their time for more nuanced, skilled input that is required for effective metadata.

It also helps if media operator teams understand why they are tagging metadata. This helps them be more connected and engaged with the workflow, empowering teams in their day-to-day work and giving them a sense of purpose.

Metadata is indeed complex: that is the nature of the beast. Therefore, it is so important to take steps to avoid metadata fatigue. Media companies must make sure that the information collected as metadata is as useful as it can be, and is recorded in such a way to remove duplication and maximise efficiency. If media organisations can streamline this process, metadata can live up to its true potential.