Big Data – no Big Easy
April 22, 2016
Broadcasters could be forgiven for wishing the digital train would occasionally slow down or at least rest a little longer in one station before moving on to another.
For many years, of course, it was very straightforward – a signal was sent, by a variety of means, to a TV where it was watched or recorded and that was it. And then the Internet came along. Not designed for TV, of course, and perhaps that’s why broadcasters were slow to embrace it, that is until new challengers made it clear that they had no choice.
The challenges of working with IP, though, remain ‘non trivial’ – especially as the train moves inexorably on to UHD 4K. But, as ever, a combination of truly impressive compression along with much less impressive increases in broadband speed, come to the rescue.
Still, while sending TV over the Internet isn’t easy, at least recovering information about the devices receiving it and, thereby ,the viewers is easy – that’s data, after all, and that’s what the Internet was invented for. So data collection, management and exploitation is the easy part of OTT, and it must be well developed and highly profitable?
Not really, it turns out. As our Big Data roundtable discussed – watch the video here – the challenges are immense and not always technological. While OTT specialists are using data to inform content buying and to deliver embryonic personalisation, and some broadcasters are at least trying to target their advertising, many more players are simply not equipped to collect the data, let alone analyse it and deploy it effectively.
The doyen of OTT and, thereby, service shaping as a consequence of data, is Netflix – who else? But its VP Product Innovation, Todd Yellin, told the BBC recently that Big Data is in its adolescence at best. ‘People think Big Data is a mountain of gold, it isn’t it’s mostly trash,’ he declared, and, naturally, the skill – in the algorithms – is in panning one from the other.
But it seems certain that the building of techniques and experience in this area will be rapid and profound in the next few years. That will bring its own challenges; data privacy is a sensitive consumer issue and regulators are rightly intent on rules that will be challenging for compliance procedures and onerous if they fail. Add to that striking the illusive balance between the viewers’ demand for free, or ‘freemium’, offerings versus the push back of the freak out factor if they start to feel stalked by ads, and one sees that while the Big Data destination is very attractive, it is likely to be a road hard travelled.
Other posts by Nick Snow: