Most of us have heard of this thought experiment: “If a tree falls in a forest and no one is around to hear it, does it make a sound?” Here’s my thought experiment: what happens if you publish a video but no one watches it?
Philosophical musings aside, video is the most effective communication technology invented in the past 500 years. But if you don’t reach your audience, your time and effort go to waste. Worse, depending on the industry, failing to reach your audience can have serious consequences. Important maintenance and safety procedures may remain unknown. New regulatory policies may go unheeded. The professor’s nuggets of wisdom may remain unlearned. Your competitor may get that new customer, instead of you.
The advent of the internet search engine in the 1990s helped us “look inside” electronic content and documents of all kinds, from static web pages to PDFs and other text-based documents. Such content could now be more easily found (ignoring, for a moment, the sheer scale of the internet). On the “content side”, tagging, linking, and other SEO techniques helped make content more discoverable. Within enterprises, when done right, the proper and systematic use of metadata ensured content was organized and easily discoverable, for example through faceted search techniques.
Video is Different
Even when adequately tagging or describing videos using metadata, videos themselves remained a bit of a black box. By its nature, video must be viewed, and understood, sequentially. Although we can “scrub” or fast-forward through video, this usually means the audio is either muted or difficult to understand. Videos are “long”, in the sense that, even if a short video contains only a handful or spoken words, those words are spread out in time. And, especially within education and enterprise communication scenarios, videos contain more and more text-based content, such as presentation slides
Looking Inside: In-Video Search
We have three approaches to looking “inside” video. In-video search makes it much more discoverable, and thus far more useful and valuable:
- Manual and automatic captioning, as with our REACH offering
- Automatic indexing of textual content inside video, for example a presentation
- The ability to ‘mark’ specific spots in a video with a title, description, and tags
These in-video search techniques make it easier to engage with content. For example, captioning, which is also an accessibility requirement in some cases, is not limited to just one language. We have customers supporting a global, multi-lingual audience using captions in up to a dozen languages. Similarly, chaptering automatically marks video to correspond with slides in a video presentation. Chaptering makes it easier for the viewer to follow along, provides context, and allows users pressed for time to quickly jump to the place they need to be.
The second, and perhaps most critical, benefit is that the same techniques that make it easier for users to engage with your video content also makes it easier to find. Captions and chapter markers, including associated text content and tags, are all indexed for in-video and global search. This unlocks the enormous quantity of valuable information contained within your video content.
User experience features also make a big difference to in-video search. You can display results in a global results set. Indicating when (in the timeline) in the video a result is found makes results from an in-video search more useful. Indicating the language(s) of the results also helps. “Deep linking” into the video starts playback a few seconds before a given term is spoken or occurs in the timeline. These allow users to search your video content just like they would any other piece of content.
Using these techniques isn’t limited to just Kaltura applications, such as Kaltura MediaSpace. As an ‘API-first’ platform, our tools can be leveraged and integrated externally as well within minimal effort.