What is the point in offering thousands of hours of content if you can’t find where it is?

watching ott vod on phone user experience credit Kaspars Grinvalds shutterstock

Content is King: billions of dollars spent on thousands of hours of programming

Source: Kaspars Grinvalds / Shutterstock

The well-worn mantra that Content is King still holds true in an industry which is spending more and more each year on a content arms race. The problem is that when so many billions of dollars are being spent on so many thousands of hours of programming, actually finding out where that King is — or at least the particular King that you are interested in — is an ever-growing issue.

This is why one of the most urgent tasks in the current OTT space lies in delivering a better user experience to the viewer. Viewers need to be able to locate that content quickly before dissatisfaction sets in and, in the new era of contract-less consumer flexibility, they head elsewhere. Churn rates for OTT services are phenomenal. OTT providers need to deliver what consumers want, when they want it, and on whatever they want to watch it on. And, as with everything else in the industry, the goal posts are moving rapidly; where once the primary function of the User Interface was to drive commerce, now it lies in delivering the personal experience

Some things are constant though. Designing the User Interface that drives that UX is not an easy task, especially given the constraints of the TV set as a delivery platform. There is a limited amount of premium screen real estate to play with, users sit an average of 3.5 metres away from the screen, and the control devices used for navigation are imperfect to say the least, and usually limited to basic direction buttons and ‘select’. All this has shaped what has become the default user interface across the industry over recent years: content is presented in rows, everything is visually driven with text kept to a minimum, the need to go down levels is minimised wherever possible, and focus state is clear and obvious (a viewer should be able to look away from the screen, look back, and still know precisely where they are). A television-based UI needs to be uncluttered, accessible, and simple to use.

What is key though in driving customer satisfaction nowadays is ensuring the ‘correct’ content is presented in those premium slots. Ideally as an OTT operator you want to give the viewer what they want to watch before they know exactly what it is themselves, surfacing the right choices from your content library so that they do not even have to scroll away from that initial home screen.

An influential late 2017 PWC survey on viewer attitudes to content discovery concluded that 55% of consumers find themselves looking for something new to watch every week and 62% struggle to find something to watch. With churn rates in the febrile US market at around 18%, giving them what they want is an important task.

“55% of consumers look for something new to watch every week but 62% struggle to find it.”

This is not a trivial task either. While the rise of the skinny bundle has grabbed many industry headlines over recent years, it is the rise of the fat channel in the OTT space that has really led to success. The US Netflix catalogue (traditionally the largest), currently lists 4124 movies and 1848 TV shows, with some estimates calculating that involves over 40,000 hours of content. And given the company’s famed use of metadata — a landmark Atlantic piece from 2014 suggested it had “76,897 unique ways to describe types of movies” and you can’t imagine that has shrunk since — each of its 139 million subscribers can potentially see a different home screen when they log on.

This personalisation is increasingly not only seen as a key differentiator between services but a de facto minimum standard too. From retail to banking and beyond, consumers are becoming increasingly used to services tailored at an individual level, and as discreet online services blend into an overall experience they are expecting the same level of service and personalisation everywhere.

The 2019 UX is all about that personalisation. Each member of a household should be able to access OTT services via their individualised home screen which surfaces different content for them and, in some cases, even has an interface which operates in a different way. This personalisation in turn is being built on increasingly sophisticated recommendation engines that take advantage of the back channel to understand precisely who has been watching what and when.

The degree of granular analysis of viewing can get remarkably precise (“To the 53 people who’ve watched A Christmas Prince every day for the past 18 days: Who hurt you?” Netflix once inadvisably tweeted) and provides a rich stew of information for content recommendation algorithms to crunch through. The ‘gimmes’ of content recommendation are still important; new episodes and new series based on previous viewings and/or transactional history are very much the low hanging fruit that still elicit impressive responses, as are genre or actor-specific suggestions. And these are being used in increasingly sophisticated ways, with different video clips and thumbnail images of the same content being shown to different viewing demographics (Netflix reckons that its artwork captures 82% of viewer focus when browsing).

AI and ML
These recommendations are only going to become more targeted too. Contextual recommendations, for example, can add in external datasets about real-time events such as the news or weather to narrow the focus even more (famous actor wins award, her movies get surfaced more frequently; Bit gloomy out? watch ‘The Fog’ tonight). Responsiveness is then tracked and this recommendation behaviour can be either emphasised or degraded in future cycles.

AI and ML are only accelerating this process too. Not only can they provide increasingly large metadata clouds for content, but they can also churn through much larger datasets in an attempt to find correlations between viewing choices and seemingly unconnected data points. Privacy issues and the GDPR are providing hard borders to this activity, but only to a limited extent. Once you simply know the viewing habits of a household, for instance, you can use AI routines developed by several companies to retro-engineer remarkably accurate segmented data about its demographic composition for targeted advertising purposes with no breach of GDPR strictures along the way.

Consumer resistance is an issue though. 60% of respondents in the PWC survey believe personal recommendations reflect shows that the OTT service is trying to promote rather than a choice tailored to their viewing habits. And Netflix faced accusations of racial targeting — strongly denied — last year when it was found to be emphasising minor black cast members in its thumbnail’s previews, claims it denied, saying that the images served were based on previous viewing history. It also added that the company had no race, gender or ethnicity data on subscribers.

Trust will be an important part of the future OTT/consumer relationship, especially as the industry builds out further data-driven models of consumption. But while 2019 UX is all about how personalisation can drive the UI, the future will be about how voice can control it.

We’re at the early days of voice still, and the impact it will have on the television UI is very much still in development. The buzzword is ‘Zero UI’, a deliberately provocative term that emphasises the move away from screens and towards Voice and even beyond to gesture control. Given the limitations of the TV remote as an input device, Voice alone holds out the promise of a much richer experience coupled with what looks like becoming a more stripped back visual design. And, as organisations become more confident in their recommendations, so they can narrow the amount they surface at any one time, allowing more room for bigger, more striking pictures.

In the end, the irony may well be that we come full circle to the early days of colour television. Four decades ago, consumers looking to find something to watch during primetime had around three or four choices, depending on what country they lived in. The goal of modern UX design is similar; to present them with three or four choices out of the many thousands available. The difference nowadays is that these will be very much the ones they want to watch not just the ones they are given. Content might very much still be King, but it’s the electoral system that chooses the monarch which is the real power behind the throne nowadays.