Media companies are using speech-to-text technology in their media asset management software to enrich their media metadata.
Data and content are often described as the lifeblood of any organization, and in the digital age it has immense value. Bill Gates articulates this well in his infamous statement ‘content is king’. What is less well understood is the value of metadata – yet it is crucial for any business looking to maximize the value of its assets. Metadata helps bring similar assets together, as well as enabling asset discovery through identification using relevant criteria.
Historically, asset management focused specifically on images and other static media. While these elements remain popular, video is now becoming increasingly popular – and valuable to organizations of all sizes. A HubSpot study revealed that 85% of businesses now use video as a core marketing tool with 60% of marketers finding that video drives more engagement for advertising. But unless a video or audio asset has an accompanying transcript, its metadata is limited.
A 2015 study by IDC found that 76% of people said digital asset management (DAM) makes it easier to find assets – reducing the time spent recreating assets that already exist but cannot be found. To cope with the increasing volume of media assets, organizations need more information to identify what makes each asset different – what each one contains, for example, keywords, themes, content, contributors and so on. It’s this information that makes assets easier to locate and makes it possible to quickly find out what is contained within an asset.
The growth of audio and video content means asset managers are now being forced to look at the tools and processes that allow them to manage all their assets effectively. Many organizations have neglected video and audio file asset management because of the legacy tools and processes they have in place. This has led to huge archives of files that lack metadata and any sort of value.
Now, with online video content growing rapidly, there has to be a focus on video assets as well as static assets. The latest media asset management software has advanced features suited to multimedia use cases – using artificial intelligence, machine learning and speech-to-text technology to extract advanced metadata information. This metadata is integral to getting the most out of both modern and legacy video and audio assets.
Metadata is responsible for powering recommendation engines and helping end users discover content that is relevant for them. Engaged audiences on over-the-top (OTT) platforms are driven through metadata on the video files. The opportunity for companies to personalize the experience for their customers is powerful. Metadata tags enable OTT services to do a better job of recommending content and syncing up users’ preferences to the metadata tags as they become richer in data.
This process is currently being done by humans through manual metadata entry. However, with the introduction of speech-to-text technology into the workflow, organizations can develop content archives and files enriched with more useful metadata. Unless a video or audio asset has an accompanying text-based transcript of its contents, it is almost uneconomical for an organization to gain any metadata value from the original file.
With enriched metadata, organizations can start to use insights to drive better customer experiences on OTT platforms. Companies can also drive better consumer engagement with content on social media and other digital channels. The power of digital communication is getting stronger all the time – and customer expectations are following a similar trajectory. Businesses need to be investing in tools and processes to ensure they stay on top of the volume of assets that need to be curated each day to satisfy customer demand.
Metadata-rich archives that can be searched easily to quickly locate the digital assets required have become crucial for any organization serious about digital asset management – and serious about business in the digital age.
Alex Fleming, Speechmatics