Our remit at Imperica is to cover a range of disciplines. The burgeoning world of "Connected TV" is one of them. Connected TV, delivering the benefits of Internet connectivity with web-like features to a TV screen, is starting to become understood by advertisers and broadcasters, and desired by audiences.
When I met Ian Valentine, Founder and Technical Director of Miniweb, it was in a meeting room at the company's offices, just off the Great West Road. Warm, friendly and highly conversational, Ian proceeded to ask about Imperica, before I asked him about Miniweb.
Formerly part of Sky's technical leadership, Valentine set Miniweb up in 2007. The Miniweb platform is very much in the spirit of connected TV, in that it provides an interface which allows viewers to discover content at programme level, rather than channel level, through content discovery - UI techniques broadly similar to web search. The Miniweb platform alslo allows for a degree of interaction which could appear on the screen in real time. It amplifies the shared experience.
Because the platform allows for this, the first challenge is to consider whether the current, generally-held definition of connected TV, such as MSN Messenger on a TV screen, is correct – and does the term justice. The concept of form following function doesn't naturally lend itself to such services on TV, in Ian's view: "The problem is that people take the paradigms of other connected devices that they use, and think that those functions are going to automatically appear on the TV. Users choose the best device for every function.” The development of connected TV is therefore the development of what we are able to make better, and to continue to align (or re-align) the matching of form with function.
A further definition which requires re-analysis is the concept of “lean back / lean forward”. This is one which has been used for a number of years within both the digital media and broadcasting industries to mean the convergence of both. As Ian explains: "The whole lean back/lean forward concept has been around for a long time - since the beginning of digital. The thing about digital is that it did two things: it produced more content through multiplexing, and to control that content you had to interact, because of the volume of channels. The second was the ability to roll out functionality that you engaged with: the concept of lean-forward. Picking up the remote, and leaning forward.”
Valentine's view is that the concept is incorrect. The ideal situation is to have the best of both: leaning back, but still engaged, without having to enter into a lean-forward mentality. Better TV is not magically produced through leaning back or forward, so the industry needs to strive towards situations where viewers are leaning back, engaged in quality content.
People had to lean forward to engage with digital content (such as with a PC), but when TV-based digital systems are developed, such as with Sky Plus, the whole goal was for people to remain leaning back. Therefore, a logical aim of connected TV has to make the environment and interface so simple that viewers do not have to lean forward, and yet still be TV.
In evolving the lean back/forward paradigm, Valentine clarifies the situations to which interaction is attractive to viewers. “We're not talking about writing tweets, updating your blog, or trading stocks on your TV. I don't think that's the killer app. What is the form and function of TV that we most want to improve? The killer app on TV is entertainment - it's what it does best. It's not the best tool to order pizza, and commercially, you make more money from porn than pizza. We have to enhance the entertainment experience, by making it connected. the goal of connected TV is to keep people in front of the screen, so subscription and advertising revenues grow, but through keeping viewers leaning back.”
A recent presentation on 10 learnings in the first 10 years of iTV, given by Valentine at the IPTV World Forum, suggests that we're really only past the first chapter in the story of connected TV. Interactive TV – red-button services and the like, are a short-term answer to a long-term problem, but fundamentals, such as the ergonomics of remote controls, need to be addressed early.
With a wider ADSL/cable install base, we now have a better infrastructure to move on. The early years of the last decade, when the base was much smaller and less people had access to home computers, allowed broadcasters to argue with the Government that iTV was good for the digital divide and social inclusion. The limited return path – 28k on initial devices – allowed for a reasonable level of sophistication for games and short-form interaction, and it also made money.
The opportunity for connected TV is two-fold: to continue the path to sustainable profits, and provide an open platform to allow the entry of new players with diverse content propositions. The challenge is to address the upscaling of technology and connectivity, without going back to the dark days of infrastructure haemorraging capital investment in networks.
“The ultimate goal for iTV was to enable Internet video. The initial thought that I had, was to provide an environment where you could deliver Internet video. My thoughts were that I go to a website, and the video is inside the website, and I must interact with that publisher, on their site, to get to the content. This is interactivity first, entertainment second.
“So, in order to enable internet video on TV, we need video publishers to build TV apps, to enable their content to work on TV. The current state of thinking is to have apps on TV - everyone's aping Apple: 'Let's invent another app store behind my TV – let's invent another technology to build apps, make it different to our competitors, so we can chew up the money and resources of major content owners so they [competitors] don't get them.' This dream of standardised TV, so anyone can publish to it, has taken 10 steps backwards.”
The closed environment of apps has, in Valentine's view, would deliver a market that is more fragmented than narrowband. The killer app is video, with the common denominator being the almost universal decoding of Internet video by TVs. Therefore, IP-delivered video can easily be distributed to connected devices, and technically, there is no reason why the device (or in an increasing number of cases, the TV itself) cannot make the TCP/IP connection, decode the video, and display it.
From the viewer's perspective, the full potential of connected TV is to be able to get to any video on the Internet, as long as it is compatible with the viewing hardware. This is not the case at the moment. “The current state of play – 'I wonder which app has got Eric Clapton in? There are only 5 apps, I'll go back to watching it on the web'. It's a hangover from the world of pizza apps, and is not connected TV.”
The constant return path that connected TV provides, is equally an opportunity for broadcasters and content producers. According to Valentine, it makes “a boring programme interesting. It suddenly becomes fun if tweets from friends are part of experience.” Programmes become vibrant and enriched through the messaging within microcommunities – the community makes it watchable, and we are already starting to see this with Twitter. It doesn't necessarily add anything to the qualititive value that the content publisher is aiming to derive, but it is a facet of connected TV that makes the programme attractive and watchable, and when managed correctly, allows publishers to get a deep level of intelligence from viewers on nuances within, and reactions to, their programming.
Connectivity also allows for further atomisation of business models. From a commercial perspective, the most important requirement of content owners is to monetise content. This is currently done in three ways: subscription; advertising; and interactivity (sponsorships, donations, pay-to-vote).
Connected TV blurs the lines between subscription and Pay TV; for example, the high level of connectivity can deliver micro-subscriptions with a richer experience than ad-funded programming over free-to-air, as is almost totally the case in UK broadcast.
The advertising model that was, until recently, the only option for UK broadcast TV was to purchase ad slots, at a weight of content and advertising largely determined by Ofcom – although Valentine feels that this balance will remain broadly the same, influenced by viewer acceptance. However, TV advertising starts to inherit a great deal from what we now know about the web. Connected TV should be able to provide additional ad overlays (delivered via broadband) which is targeted to households, and sold on a different model – such as CPM or CPC basis. Clickthroughs would be able to display mainstream TV advertising that has finished its broadcast airtime, which brings the long tail into media planning, as well as the advertising revenue stream. This could massively increase the longevity of ads, as they become more targeted, and shown to target audiences for longer periods of time.
The inventory will be capable of some interaction, but Valentine feels that it will largely consist of pure video. He believes that viewers will not want to use TV advertising to undertake complex transactions, but an impulse-response is fine, such as submitting details for a brochure, as long as details are stored to avoid retyping. It can also provide better integration with two-screen environments, such as working with 2-way SMS. The overall impact is a positive one for TV: more revenue, more viewed hours.
Broadband to the home allows for a much wider potential in terms of accessible content, starting at recently-broadcast catch-up content, then working backwards through broadcaster archives. If broadcasters with catch-up services increase linear viewing, then the when people want to watch, is as important as the where. What will broadcasters do, if their channels do not have something of interest to the viewer right now? In a connected TV environment, viewers could go online from the channel, and watch content down the wire. This seamless integration of broadcast and broadband experiences needs to be totally seamless, to keep the viewer within the branded environment. And, if viewers stay engaged, Valentine considers that revenue can be generated from them across both broadcast and broadband. Once viewers “disconnect” from the TV, even when using a laptop or tablet while in front of the screen, they are lost. The level of connectivity helps to surface both the range and depth of content, but its economics are different. Clearly, it can help to surface niche and specialist content, as long as the business model for viewers to discover it, and for systems to distribute it, is less expensive than over the air. Valentine sees this as potentially creating an explosion in both new and hard-to-find archived content. This, once more, could increase viewing time, at a cheaper rate to deliver content to screens than is currently the case.
All of this inevitably changes the relationship between TV (and what TV is), brands, and advertisers. A channel is effectively a branded package of content. Viewers have a branded affininty with Dave, Sky One, Sky Sports, and so on, because of their particular kind (or mix) of content. These broadcasters will then produce or buy content based on their brand personality, then sell advertising based on this brand value. “That's what all channels are. Packaging mechanisms to market content, using particular mechanisms, to segmented audiences.”
However, this is where things change. In a world where a lot of content is online, a lot of channels disappear. +1 channels disappear if the content broadcast on +1 channels are readily available. Connected TV will allow for small-audience reruns, as viewers will be able to discover more content, and through an increased range of discovery tools. Niche over-the-air channels will disappear according to Valentine, as online becomes the prevalent means of distribution, leaving the core broadcast channels showing live sports, live news, appointment-to-view and premium content.
The concept of branded channels then completely transforms. It is a new answer to an often-asked question: are viewers looking for a channel or content on EPGs? Once online, Valentine considers that viewers are not thinking about the publisher brand; they are thinking about content, which represents a shift in viewer discovery to content, rather than the packaging. The impact is that lot more small, specialist content becomes available, because discovery tools take you straight to what you want; as with the transformation of the music business, content producers no longer have to be part of a major brand to get carriage.
While channels are downplayed in the light of increased content discovery, connected TV also introduces a new role. Services around the discovery of content shifts from brand packager – the channel - to service providers. Valentine considers this to be a shift that has already taken place on the web: “Consider what happened on the Internet, with portals such as AOL, Yahoo: it ended up with Google having the value. So, it won't be BBC iPlayer, ITV Player or Hulu - but the service brand that gives you the full, solid connected TV experience, independent of content owners”.
Service providers will also play the role of concierge, undertaking administrative-yet-meaningful roles such as storing viewer details in the earlier example of interactive advertising. If the role of service provider starts to become meaningful, and we are moving to a world of content discovery, I question Valentine as to whether many in the broadcast industry don't “get it”.
“Absolutely. That is the state of the industry right now.” He cites iPlayer as a case in point; that the BBC actively markets it, and try to keep users within it for as long as possible. Each app is therefore its own silo of content: the app's services are only available to viewers within it. It might suit content owners very well, but Valentine makes the point that the “endgame of any tech is to strike a balance of statisfying requirements for all parties involved... but, you have to think about not just the requirements of the publishers, but the requirements of manufacturers and viewers.” Valentine doesn't consider that these views are fully addressed at the moment. “with the 'five apps on a connected TV' model, because all of the manufacturers have the same five apps, and once you're in the apps, all of the functionality is exactly the same. Where's the differentiation? They want to be in the value chain and provide something special, but they don't know what it is yet. As far as the viewer is concerned, the experience sucks. I'm in and out of apps; I have to play content which stops, and I'm back in the app. I'm leaning forward all the time. When I watch TV, I expect the entertainment to come to me.”
Manufacturers and viewers must have the “brains” to determine what to watch, as they cannot be held within a single content provider, if there is to be a shift in power from broadcast channels to content. This same shift also needs to make online content more of an integrated experience for viewers.
“Do we think that paying for content is going to work the same way as the web? This is the web experience: I see a trailer on YouTube that looks interesting. Baked into the trailer is a URL. I write the URL down on a piece of paper, and type it into the address bar. I then arrive at the website that has the real content, that the trailer was for on Youtube. I want to see the content that's paid for. So, the website is asking me to register. I have to type in a credit card number, get a login, get my email verified, and then maybe I can become a member of their website, so I can see that piece of content that I thought that I might want to see because I saw their trailer on YouTube. I'm setting up another payment agreement with another website. That's the best experience on the web, because YouTube would never link through to a piece of premium-rate content.
“Now, think about this on TV. I watch a trailer, and it is a trailer for a piece of long-form content. Wouldn't it be great if the connected TV service could understand the relationship between the publisher of the trailer, and the publisher of the long-form content, and say 'Now you've watched the trailer, would you like to see the long-form content?' You want to see it, and it responds to say that it already has your payment details, you can watch it for 50p, or everything from this provider for £9.99, for the next month. I go straight into the content, but we've improved the Internet video experience on TV over the web, and that gives the viewer a reason to stay there. Those improvements don't come from following the web model, with content locked into sites; those improvements only come when a manufacturer or service provider can provide discovery, personalisation, payment services, and metadata relationship services that span all of the content providers. It benefits the viewer. It's inevitable.”
The sum total becomes an added-value proposition for advertisers, who benefit from a greater degree of knowledge about who is viewing what.
The basic ingredient to make this change is data. In connected TV, you now know much more about the viewer. In broadcast TV, you don't know anything about viewers – they are all the same. In the Internet you know a little about the viewer – you know a little bit about them through their cookie history, bots, and IP georeferencing. It's a little more targeted, and of course you can contextually place ads around content.
"Imagine a connected TV device that knows everything about you, what you've watched, and can serve dynamic advertising, even down to an audience of one. There is huge data in such a service, and it is of value to advertisers. It should be better advertising for the viewer, almost enjoyable and useful advertising. It should be much more valuable to the advertiser, because they're not wasting money tracking it. TV's best advertising destination is video, which is the best way to sell yourself. Even on the web, the highest-value advertising collateral is video.”
The value of connected TV advertising is such that multi-screen retargeting also becomes a true possibility. Again, we return to form and function, and the basic principles of how different devices work. It's important is to synchronise the user profile in the cloud across devices, and to have some kind of user profile on TV without the pain – and unfamiliar concept – of logging into the TV. Connected TV software and platforms need to work out who is watching, with Valentine suggesting that the provision of individual recommendations, bookmarking and community, makes it enough of an offer for viewers to indicate who they are.
The interconnectivity will therefore be central to the development of a truly synchronised, cross-media, user-focussed and highly contextualised advertising offer. It is universally accepted that the phone will play a critical role in any interrelationship, with two-way text being part of the general level of interactivity which spans both advertising as well as the content itself. Having the power to influence what is happening in the content, or have an extended level of interaction on the phone coming out of the TV content, could make live or near-live multimedia storytelling an even more interesting, dynamic proposition than is currently the case.
We finally talk about the state of the broadcast and production market in the UK. For connected TV to work, and mass content to open up to a mass viewing public, there needs to be a greater dialogue between Internet publishers and TV broadcasters. What do they need to understand from each other to make it all work seamlessly?
“They are different beasts at the moment, but there is overlap. There is a change in the business models. Internet publishers need to learn that there is revenue in subscriptions to premium-rate content on TV. On the web, everyone assumes it to be free. On TV, people want a good entertainment experience, and are willing to pay for it. So, web publishers can learn that some of their content is valuable enough to charge for, and if there's an easy charging model, that's a valid revenue stream that is could be part of their business model, as well as pure ad-funded content.
“Broadcasters need to learn that there are Internet advertiser models, that are applicable to their audience. If you imagine the advertising model around a blog – based on eyeballs – think about how many eyeballs hit a TV channel every day. There are huge Internet-style ad models that can work around broadcast TV. Then, it's content. The skillsets of creating good quality entertainment need to flow out to a wider range of publishers. Geekbrief TV is still one of the most-watched podcasts; the production quality is good, and it's short and punchy. Broadcasters need to embrace online publishing. Opening up the whole of your archive and catch-up content is going to be huge, and they need to learn not to fear that. They could do that, or pass it to aggregators such as SeeSaw.
“Soon, we'll be able to watch anything, anywhere. That's really the ultimate endgame: to have all of the content on all of the devices. If we start thinking about the industry from that point back, we can see what's going to have to change.”
Ian Valentine is Founder of, and Chief Architect at, Miniweb.