We’re all broadcasters now – if we want to be.

The revolution has happened and now anyone with the right connected device can create and distribute content. The number of connected devices has exploded over the last five years;Ericsson’s Media Vision 2020 predicts that there will be over 50 billion connected devices by 2020 – that is literally billions of new places to watch your favourite show – eight billion of which will be mobile broadband subscriptions. The result is that many more people in all sorts of new locations will have access to content with none of the old boundaries; and the flexibility of mobile network coverage is giving more people the chance to access prime content than linear broadcast ever could. Regional release dates for content have squeezed together or even collided as consumers have come to expect access to anything, anywhere, anytime and preferably as soon as it pops up within their social network.

Consider this; you’re on the train and you remember a great TV comedy moment you want to share with your friends. You can remember the name of the series, the name of the actor; you’re even word-perfect on the line itself. You tap in the search details, start trawling through the list of suggestions, but, as it stands, it’s going to be a matter of luck and a little cyber-sleuthing to track down the right episode – and even then you’ve got to spin through the whole thing to get to the line you want. It’s much easier to be able to do a text search on the script and go straight to that point in the scene instantly – thanks to the captions.

What you may (or may not) know, is that the essence of a caption is a piece of timed text, and new production processes mean that we can now add even more granularity with word-level timing. This means that through good quality captions and associated metadata you can make it much easier for consumers to find your content.

If any user anywhere can find your piece of content (or more importantly the right part of it) then that creates many new commercial opportunities; but it needn’t stop there. With the right processing of the words by those skilled at extrapolating and collating data concerning that clip, it’s possible to create a rich set of metadata that can enable even more abstract searches, links and presentation features. These days it’s possible to identify the shot changes, the scene boundaries, the divisions between the speakers (and link them back to that character or actor); captioners can even create bespoke metadata like key events (like the goals in a football match) during the preparation process.

Natural Language Processing (i.e. software or systems that can ‘understand’ human speech) can assist in the creation of automatic keyword data – what were the important words used in that clip, and what other clips have a similar profile? If you know this, you can start to link content on new levels – rather than being limited solely to the “if-you-watched- this-you-might-like that” kind of recommendation (which I find is often more of an insight into the strange nature of others’ viewing habits rather than your own).

If you’re able to profile the content and the appearance of themes and phrases, you can start to make a link to brands that might like to be associated with that content. If we know the timeline of these themes as they change across the clip, then it’s possible to create advertising or marketing opportunities that match the viewing experience. A logical step on from this is to spin it around – and make sure that you never show the wrong advert next to the wrong pictures.

Finally, what about using caption data to extend geographical reach, breaking down the language barriers with a real-time transcript? With Machine Translation reaching new levels of accuracy (Google, Microsoft et al have demonstrated new cloud products in this space – with pretty good levels of intelligibility), it should be possible to create an online translation of a captioned clip on the fly, into any language the viewer chooses, and present it back as captions. Cutting edge comedy will always present a challenge for such systems (indeed it could introduce its own variant of surreal humour) but it will certainly open up content to an ever larger audience with little extra effort.

Captions are not limited solely to accessibility – they can also help bring viewers to your content, and generate more money from it.

Matt Simpson, Head of Technology & Strategic Product Management, Access Services.