How Personalization Can Help Drive Pay-TV Viewer Engagement

Mon, 22 Feb 2016 | By


One of the most debated subjects at industry trades shows in recent years and one of the most significant factors deciding the future direction of the industry is personalization. Personalization, often now powered by the cloud in a suite of technologies as TV as Service models are established, drives and deepens engagement, and engagement is, without a doubt, one of the goals for the current industry that can be labelled with the sobriquet “holy grail.”


Indeed, a recent survey from The Diffusion Group, The Recommendation Revolution “ The Future of Recommendation-Driven Guides, 2016-2025, estimates that recommendation-based guides will be driving 75% of all TV viewing by 2025.


Here’s what Rapid TV News had to say on the matter: According to the study, US multichannel video programming distributors (MVPDs) will rapidly move to make recommendation-based program guides the centerpiece of their updated TV interfaces. This momentum is so strong, TDG suggests, that within the next decade, the use of these recommendation-driven guides will flourish, ultimately driving three-quarters of all TV viewing


Personalization Factors


There are several factors in play here that decide why this is such an issue as we move forwards into the cloud-based TV future of 2016. Firstly, the landscape is increasingly fractured: what was once a simple delineation between traditional broadcast and IPTV is becoming increasingly blurred, with audiences chopping and changing between Pay-TV, VOD, SVOD and myriad other services (we would be remiss if we didn’t, unfortunately, include piracy in that mix) almost at whim. TV as a Service will only accelerate that churn.


Engaging with them, maximising loyalty in the face of the increased threats of cord-cutting and cord-shaving is the key to future business success.


Secondly, we have to look at the sheer growth and depth of content available to people in the modern world, which even if we simply narrow the focus down to television is still running at a level that would have been undreamt of several decades ago. There are 1000s of channels showing 1000s of programmes, a huge torrent of data and content that is beamed around the world by satellite and transmitter while further exabytes of it are carried by the internet which nowadays is mainly comprised of cloud-based video signal applications rather than information.


To paraphrase the old internet activist adage: information wants to be free but there’s video to watch (and that costs money).


Thirdly the modern consumer is a proactive beast. As a society we have become used to mining information resources, saying goodbye to any pretence of brand monogamy in a constant churning process of measuring goods and services to get the best prices and deals for ourselves, aided and abetted by increasing competition between the providers that are vying for our dollars.


The challenge facing the pay-TV market is weighing all these considerations up, mapping the fractured landscape, and finding a way for the newly proactive consumer to navigate through the sheer volume of content out there to their own services.


Happily, and with the aid of new technologies like the always-on, on-tap power of cloud computing, that can be done, and one of the best ways of doing that is via personalization.


Personalization and interaction


Proactive personalization and contextual interaction, often powered by TV services in the cloud, is undoubtedly one of the best ways of engaging with viewers.


And yes, the key word there is context. A fully contextual user experience is where systems and devices begin to learn and adapt to our needs, and to a great extent start to predict them. They have to be implemented carefully, since consumers never like to feel that they are being led via the nose and need to retain some sense of agency in their own actions but personalization, in this sense, is a powerful tool.


There are four different ways that the algorithms powering the recommendation engines that sit behind this effort all work: collaborative matches the user’s consumption patterns with similar users; content-based relies on a user’s ratings; demographic matches content from similar demographic profiles; and knowledge-based extrapolates from provided preferences.


While all of them are becoming more powerful thanks to being able to leverage cold computing power, all of them also have their strengths and all of them have weaknesses. These are algorithms trying to mimic human behaviour and that is as variable as the seasons. Which is precisely why there is also a very hefty Big Data effort involved in all this activity too as analytics tools try to mine what we want to watch and why.


This is where contextual information comes in and comes in at a level that, when it’s right, it almost comes across as magic, TV as a Service is that powerful. The idea is to blend user-provided information, which is often freely given if the provider has done a good job of incentivising the information gathering process, with more obvious data such as what time of day it is, what day of the week it is and what season; information provided by a device such as what they’re watching on, geo-location (this can be a really interesting source of contextual information) and even if the user also has friends in the vicinity); and the techniques can even be spun out to datasets that might seem fairly esoteric to most people (weather forecasts, sports results and so on).


Much of this information is already available in the cloud. All that is required is to stitch it together in a coherent manner


All of this information affects what we want to watch and when we want to watch. Gathering enough data and being clever at the way that a system crunches the numbers can give the best contextual recommendation engines a fairly good percentage chance of figuring out precisely what that might be.


It is estimated that Netflix has amassed more than 77,000 unique ways to describe movies. And if Netflix can do it, it’s less a matter of that anyone can, but that their rivals probably need to.


A Personalized Future


We are at the dawn of a future of smart personalization in many areas of our lives, aided by the growth in the cloud and the introduction of TV services in the cloud, and for the broadcast industry times have changed significantly. In many ways it has almost been a complete reversal over the decades, from a past where a handful of program commissioners and schedulers decided what the audience would watch and when, to one where the audience tells the content producers and providers what they want to watch and simply go elsewhere if it is not provided.


As to when they want to watch it: that answer is simple. Whenever. And on whatever device they want to. In many ways, the viewer and the TV service are now both in the cloud.


Television needs to adapt to the new requirements. The future of many industries is less about selling products and more about managing customer journeys and expectations, and providing effective personalization is a way of accelerating that process, providing short-cuts, and ensuring that an engaged consumer that is being guided to what they might want to watch is a loyal one too.