One of the most daunting technological challenges we face today is scaling up this old internet of ours to meet the burgeoning consumer demand for bandwidth-intensive real-time applications such as telecommuting, cloud computing, and streaming media.

And as internet video continues to trend from short-form/long-tail/low quality content towards long-form/short-tail/high quality (premium) content (i.e. from YouTube to hulu to TV/films on embedded hardware), exploding consumer demand could bring things to a head even more quickly than currently anticipated.
Read the rest of this entry »

Televisions and set-top boxes with embedded network interfaces are coming, that much is agreed upon.  What’s up for discussion, though, is just what the user interface is going to look like – an “internet video-only” implementation that places a premium on simplicity and system stability, or a full-featured “browser for the couch” allowing full unrestricted access to the internet.

Gordon Campbell, a 30+ year veteran of the semiconductor design and marketing industry with stints at Honeywell, Motorola, Intel, and several start-ups under his belt, calls the former approach “hogwash.”   According to a recent article, his current company Personal Web Systems (no web site yet)) plans to bring a device to market later this year allowing full unrestricted access to the web (the company also has plans to subsequently offer that same functionality to CE manufacturers on a single chip).

“This generation doesn’t want their hands tied behind their backs. They want the same experience as with a PC (on their TV)“, Campbell states.

I think he is precisely wrong.

Television web browsing has been tried many times before, with little or no success – and although today’s increased broadband penetration and (more text-friendly) HD screen resolutions suggest perhaps it’s time again to make yet another pass at it, the bottom line is that average folks just do not want a lean-forward PC experience on their TV, thank you very much.  And even if they did, there would be user input device issues to solve (keyboard on your coffee table, anyone?), challenging security issues to deal with,  and (in contrast with computer users), zero tolerance for crashes and restarts.

I could go on and on – but in short, I feel it’s a mistake to assume that internet access necessarily dictates a full PC/web browser paradigm – for example, consider twitter, skype, IM, even iTunes – all examples of succesful non-browser/non-PC dependent internet applications, none of which “this generation” would consider “hogwash”.

Once you explain to them what the word “hogwash” means, that is.  

To sum up, the consumer electronics industry has discovered the internet – and these new devices are not going to need (or look anything like) a PC or a web browser.

For better or worse, the TCP/IP protocol (and the internet it makes possible) will not remain the exclusive turf of the computer industry for very much longer – a point some in the industry are slow to see.

We’ve written before on how ill-suited the personal computer is for viewing long-form internet video – and on the strange inability one often finds in the personal computing industry (and in a lot of new media analysis) to distinguish between how a 2 minute YouTube video and last week’s full episode of Lost are actually consumed.  The point is not lost on the CE industry, though: there will be a deluge of internet-enabled video hardware coming to market within the next 6-9 months (both televisions and set-top boxes), and while the few devices already out there (i.e. Apple TV, Vudu, and Roku) have all been based on closed “walled-garden” models, this new generation of hardware will instead be open, offering the promise of access to multiple internet video sources directly from the couch.

Which begs the question: what should the user interface for a system that aggregates multiple (and often competing) video services look like?  Clearly, a wide-open web browser model isn’t the appropriate solution for what is, after all, a consumer electronics device.

From an application design perspective, it’s an interesting question. Although I’ve already written about the approach Yahoo/Intel are taking with their Connected TV initiative, last week I had the opportunity to speak with Bob Poniatowski of TiVo regarding their upcoming internet video solution (currently in beta testing).  Two things I took away from our chat: (1) TiVo continues to place a substantial premium on UI design and ease of use, and (2) they’ve determined that focusing on a searching (rather than browsing) model neatly solves the problem of how to integrate multiple internet video services into a single cohesive user experience.  In fact, the name of the initiative (to be rolled out later this year as an additional feature on existing Series 3 and HD boxes) is “TiVo Search” – as CEO Tom Rogers puts it, “what Google did for the Internet, TiVo is now doing for the TV”.

It’s all about the search: users will be able to look for short-form content from sources such as YouTube, The N.Y. Times, and The Onion (among others).  As for premium content, if you have an account  with Amazon VOD, CinemaNow, or Netflix, you’ll enter a TiVo PIN on the respective website and be good to go.  However, one caveat: searching on Netflix is not yet supported – like the  Roku device, only whatever “Watch Instantly” titles already added to the Netflix queue via their website are available.

As an example, search “No Country for Old Men”, and you’ll be able to compare, purchase, and view the title from either Amazon or CinemaNow if you have accounts there (TiVo transparently handles any transactions).  You’ll also get reviews and related articles (from the N.Y. Times, for example), and from Youtube, you’ll get trailers, clips and fan raves/rants (Poniatowski likens the YouTube content to that of a “global DVD Extras menu”).  Search Tommy Lee Jones and you’ll get bio information, any other available films and/or television programs he’s appeared in, and again, any related short-form and user-generated content.

In addition, TiVo Search will include a (very TiVo-like) internet video “Discovery Bar” of suggestions based on your previous searches, and will also allow you view images from any computers on your home network… all in all, it’s easy to imagine this being pretty cool.

Things to watch:

• How will TiVo’s subscription revenue model compare to Yahoo/Intel’s Connected TV advertising-supported model?
• How will TiVo’s traditional in-house software development/deployment model compare against the Yahoo/Intel Connected TV “widget” model (and/or Apple TV’s App Store model)?
• When will Netflix “Watch Instantly” content become searchable too?
• Will TiVo expand into the lower end of the IPTV market by releasing a more affordable streaming-only (no HD, no DVR) device to compete with devices such as the Roku?  Having already done the heavy lifting of implementing the search system together, this would seem an likely move.

All in all, this looks to be a powerful and (as one would expect from TiVo) a well-designed long-form internet video solution.  Although TiVo’s market share has been under pressure from lower cost carrier-bundled DVRs in recent years, TiVo Search could be just the differentiating value-add the company’s looking for.

I’m looking over some recent numbers from nielsen online and I’m struck by something: while hulu may indeed be a big fish, long-form internet video is still a pretty small pond.  Consider:

Of the four major broadcast networks, hulu partners Fox and NBC saw the largest month-over-month increases in October 2008:  Fox was up 165%, while NBC (helped by Tina Fey’s triumphant if temporary return to SNL as America’s favorite eye-winking, Russia-seeing hockey mom) saw a whopping 312% increase (by contrast, ABC was up 105% and internet video laggard CBS was only up 38%.)

312 % and 165 % increases over the course of one month? Let’s celebrate – professionally produced long-form video has finally come into its own, right?

Wrong.

From that same Nielsen report, here’s another statistic: during October, YouTube had almost 82 million unique visitors to hulu’s 6.3 million – that’s a factor of fifteen (even with Tina Fey’s Palin sketches driving users to hulu).

A direct comparison between the two by total streams delivered would skew unfairly towards YouTube due to the shorter running time of the average user-generated video – but what the heck, let’s do it anyway, just for fun…  because the difference between those Nielsen numbers is even more stark than you might imagine: YouTube delivered almost 38 times the total number of streams delivered by hulu.

That’s thirty eight times more streams from YouTube than hulu.

Granted, hulu is one well-executed website.  Yet clearly, long-form premium video over the internet still has a long way to go.  What’s the takeaway here?  In my opinion, the answer is somewhat obvious: people don’t want to sit alone in front of their computers for a half hour or more at a time to view long-form video – in other words, the effectiveness of the personal computer as a video-viewing device is inversely proportional to the program length of the video being viewed.

The numbers in this report clearly put Hulu and YouTube in stark contrast against one another in terms of actual usage.  However, it would be a mistake to fail to take into account the fundamental differences between the short form/long-tail (user generated) and long form/short-tail (professionally produced) video viewing experiences – or the fact that we don’t have a truly compelling lean-back device for delivering long-form internet video viewing just yet.  Therefore, it would be a mistake to infer from reports such as this that internet video will remain primarily a short-form UGC medium.

For long-form premium video over the internet, it’s going to take a new generation of device that offers content directly from the couch before we can make any such comparisons.   The user interface on these devices will not be a web-browser, instead it will be simpler and optimized for lean-back media. Companies such as boxee (at left) and Yahoo/Intel are working on just such user interfaces. While I’ve already written a bit on the Yahoo initiative here, Boxee is more recent development. Right now it’s just a Windows/Mac application that aggregates disparate video sources (including Hulu) into a cohesive whole. While that’s pretty cool in itself, what makes Boxee really interesting is that the company plans to bring dedicated set-top Boxee hardware to the market within the next year or so – and in ther meantime, the software can be installed on the Apple TV device today. As I’ve said before, I think the prospect of Boxee on – well, a box – changes everything.

We shall see – but in the meantime, a quick reality check is in order: while well-suited to workplace video snacking, the computer and web browser are inappropriate (and ultimately intermediate) solutions for viewing long-form video – no matter how well-implemented a given website (such as Hulu) happens to be.

Consider Sony and Apple – in many ways, two similar companies: both span both the computing and CE spaces, both have intimate connections to a major studio (Sony Pictures and Disney, respectively), and both adhere to closed-end vertical silo business models.  Granted, Sony doesn’t write their own operating systems, and unlike Apple TV, the current Sony internet-video-to-the-television box (the Internet Video Link) is partnered with 3rd party services such as Amazon’s Video on Demand – but in other ways Sony represents more of a closed ecosystem than Apple: while the Apple TV will work with any HDMI-equipped TV, the Video Link will only work with Sony Bravia televisions – and while iTunes is platform-agnostic, Sony’s previous ill-fated internet TV device from a few years back (the Sony Room Link) demanded not only a PC, but a Sony Viao PC.

This past week saw some news from both companies:

• Sony announced an expected $US 2.9b operating loss for 2008
• Apple recorded a year-over-year revenue increase of over 6% for the most recent quarter – and this despite the historically horrendous macroeconomic climate of the past few months



It’s an ongoing debate among those of us who think about consumer electronics and technology: closed proprietary platforms vs. open standards-based platforms.  Stability and elegance on one hand, lower costs and increased innovation on the other – two entirely different paradigms.

In addition to their numerous other circumstantial similarities, Sony and Apple both subscribe to the former – so maybe it’s not about the intrinsic advantages of a closed or open technology model (or other factors, for that matter) as much as it’s about the quality and desirability of the product.

The Holy Grail: internet video directly from the TV.

While the primary challenges would seem to be hardware-based,  there are equally significant (and equally daunting) software challenges to be met as well, because no computer = no OS = no web browser.   Traditional operating systems aren’t appropriate for a lean-back passive viewing system on any number of levels, and studies repeatedly show that nobody really wants a full-featured browser on their television (and a keyboard on their coffee table) anyway, so in one sense it’s no great loss.  However, forgoing all that standardized functionality leaves one with a lot of design and development work to do: what about an operating system?  What should the user interface(s) look like?  Can we get by with the current button-laden remote as an input device, or should a Wii-like pointing remote be developed?  And how does the type of input device dictate what type of functionality can get built into the system itself?  All in all, a pretty heavy lift – because what we’re talking about here is the creation of an entirely new interface to the internet (potentially every bit as important as the web browser).

Into the void step Yahoo and Intel.  Just announced at last week’s CES is the Connected TV Widgets Channel, a ‘software foundation’ for internet-enabled television hardware built around a new generation of specialized Intel processors.  While the framework will be open to 3rd party software developers (according to Yahoo’s Patrick Barry, “We get a nice advantage, knowing the ins and outs, but we will not limit the platform to being addressable by us”), it’s worth noting that Yahoo and Intel are going with a lightweight “widget” model rather than a heavier-weight application model.   Running a widget on a modified JavaScript engine rather than an installed application down on the operating system itself tends to protect the OS from poorly-behaved software and also allows for more generalized control of what the software can and can’t do.  Much like the Apple App Store model, the widget model represents an attempt to strike a balance between encouraging open and innovative software development, while at the same time providing appropriate “guard rails” for what are, after all, consumer electronics devices rather than computers (look for this trend to continue as cloud computing and the overall “CE-ization” of home computing continues).

Initially at least, the Widget Channel appears to be primarily about adding ancillary features on top of the traditional cable/satellite television you already have – in other words, imagine being able to call up feeds from fan message boards or team websites in a dock at the bottom of your TV screen while simultaneously watching the big game on cable…   Or having a dashboard of specialized weather, news, or twitter feeds available while watching “Madmen” via satellite dish…

• On a purely technical level, though, there’s nothing to prevent a Connected TV widget from streaming video, either (bandwidth permitting).  At that point, things get interesting – as the innocent little ‘widget’ starts to eat into existing television distribution models.

• In fact, the terms “Widget” and “Channel” are both misleading, because the Intel/Yahoo initiative is not about merely adding additional incidental functionality -  it’s about (cue the thunderclap and the dry ice) letting internet video into your television.  In other words, that local weather report or eBay quote on the bottom of your screen is really something of a Trojan Horse (a point already probably not lost on the cable industry).

Connected TV is, um, well-connected: on the hardware side, CE manufacturers such as Samsung, Sony, and LG are already on board, and for web content, deals have been inked with traditional heavyweights such as eBay and the New York Times (among others).  For the video over IP scenario to play out, though, what Connected TV needs are video content partnerships – and there too, Yahoo and Intel seem to have things well in hand: agreements have already been signed with CBS, Netflix, Amazon, Blockbuster, and Showtime

It’s been a while since we’ve seen much good news coming out of Yahoo, but they seem to be getting a lot of things right here.

Yahoo intends to monetize the Widgets Channel through advertising, but in an effort to reach a critical mass of users as quickly as possible, will reportedly go easy on the advertising initially.  So who knows, maybe in a few years from now, Yahoo stockholders could actually be thanking Mr. Yang for turning Mr. Ballmer down at $31 per share….

It was almost two years ago now that Apple dropped the ‘Computer’ from its corporate name, reflecting the company’s growing presence in the consumer electronics space.  We believe the next iPod-like disruptive CE devices will be an completely new generation of network-enabled TV and set-top hardware offering direct and simple access to internet video ‘baked-in’ – and we believe some compelling implementations are right around the corner.

Meanwhile, consider an exchange from a panel at last week’s ‘Future of Television East‘ conference in NY: asked whether the emergence of such hardware would have much impact on the (currently computer-centric) internet video space, a panel member representing a major software firm essentially reduced the issue to that of merely a difference in user interface ergonomics  – keyboard/mouse vs. remote.  To paraphrase the panel member, “the user will get to the content whatever way the user gets to the content” – but yet minutes later, this same panel member was admitting that user uptake of their internet video TV technology was “not happening as quickly as we’d all like to see”.

Why is that?  Well, if a company’s solution to the problem of getting video from the internet to the TV involves putting a personal computer between the television and the internet and administering it all from the desktop rather than from the couch, they are just not going to see much uptake from Joe the Plumber – instead, the average user will wait for a ‘lean back’ solution to become available, saving him or her from having to get up and touch the computer at all.  Although this is precisely the point at which we currently find ourselves, we don’t see the situation lasting very much longer – again, we believe internet–enabled TV hardware will be the big story of 2009 (one of the issues that’s kept this from happening so far is the development of a useable motion/pointer remote, but that’s on its way too – more on that here).

Clearly, there are major ease-of-use differences between the personal computer/media server internet TV model and the (as yet unavailable) embedded hardware internet TV model – in terms of set-up, user experience, and maintenance.  But while the various software incumbents might have a stake in downplaying these differences, they fail to recognize the inevitability of standalone internet video devices at their own peril – because this new generation of hardware will also require a new generation of lightweight system software and highly efficient video codecs/players (a potentially substantial new market).  Furthermore, to the extent internet-enabled TV and set-top hardware is successful in the broader mass market, a new user application market could also emerge for these ‘quasi-computers’ (Apple’s App Store is a good example of an appropriately simple and elegant approach to distribution and administration for such an entirely new software model).

Amazon has an industry-leading online retail business to support its ‘Video on Demand’ internet video initiative. Netflix has an industry-leading DVD rental business to support its ‘Watch Instantly’ internet video initiative.  The iTunes video service?  Apple is, well, Apple.   And Movielink is owned (and supported) by Blockbuster.  Alone among the major video services, relative old-timer CinemaNow was out there by itself in the still very nascent long-form premium content internet video space.

Until yesterday.

As Sonic Solutions’ Chief Technologist Jim Taylor was sitting on a Future of Television East panel here in New York yesterday, his company was announcing the acquisition of CinemaNow.   Known for the Roxio DVD authoring software, Sonic had recently chosen CinemaNow as the storefront service partner for their Qflix DVD-burning system (CinemaNow’s other main partner to date is HP, which offers direct access to CinemaNow from their MediaSmart line of HD TVs and Connect set-top box – in fact, both ship with $20 CinemaNow coupons).

Since we believe the next phase of internet video will be about long-form premium content accessed directly from dedicated  network-enabled CE hardware, we were particularly interested to read Sonic Solutions’ CEO David Cook’s take on the move:  “With broadband-connected consumer electronics hitting the market in ever greater numbers, there is a growing need for a service that gives consumers one-click access to premium entertainment on any device in the digital home.  The combination of CinemaNow’s content and embedded device strategy with Sonic’s technical prowess and broad PC and CE distribution promises to fulfill CinemaNow’s original mission.”

A smart move for Sonic Solutions…  unlike the current Qflix system (which requires a PC), look for a combined Sonic/CinemaNow to offer Qflix-enabled hardware with embedded access to CinemaNow – no computer needed.   Such a device could compare very favorably to an AppleTV or Amazon/TiVo solution.

I’ve long been of the opinion that longer-form (professionally produced) internet video will happen on a truly widespread scale only when the problem of getting that content over to the television is solved – and that current discussion of internet video (and how to how to best monetize it) is often based on two questionable premises:

1. The personal computer will continue to be the primary internet video delivery device.
2. Internet video is necessarily about short- and mid-form content.

Such discussions often completely fail to adequately recognize how profoundly game-changing direct access to the internet from the next generation of TVs and set-top boxes will be.  In other words, while short form internet video (user-generated or otherwise) will always be a workplace diversion, the main event has not happened yet – we’re still in merely a transitional, evolutionary phase of the process – a process which will end at the couch, not the desk.

Where are we now?   Several major CE manufacturers are already offering a first generation of standalone internet-enabled devices (each partnered with one or more internet video services):

Hardware:                                Service:

• Sony Bravia, PS3               Sony Playstation Network, Amazon on Demand
• Roku                                   Netflix, more to come…
• LG                                      Netflix, more to come…
• AppleTV                             iTunes
• TiVo                                   Amazon, Netflix
• HP MediaSmart                  CinemaNow, including others
• Microsoft xBox360            xBox Live Marketplace, Netflix

The problem with the above scenario is that no computer means no web browser, which in turn means no Flash – so each OEM wishing to offer multiple services directly via their network interface-enabled hardware (TV, DVD, PCR or set-top box) has had to implement the interface to each service partner individually – a tremendously inefficient reinventing of the wheel.

Therefore, I’ve long felt that what’s needed is a standardized technical protocol for CE hardware to interface with these multiple video services.  Last week, Reed Hastings of Neflix expressed a similar view, noting that in the absence of standardization, “Everyone’s going to have to do customer interfaces for each device”, and further, that it’s “slowing down the market tremendously.”

While the web standards we now take for granted were developed in the shelter of academic and government agency environments, there’s a huge amount of money (and an equal amount of competing agendas) at stake in the internet video space – so the development of a new standard from the ground up at this point seems highly unlikely.  Instead, an embedded web browser running Flash (and/or Microsoft’s Silverlight) sounds like the better idea (Sony has already moved in this direction, embedding the highly-regarded Opera browser into its Bravia line of network-enabled TVs).

Admittedly, though, web on the TV leaves a bad taste in the mouth – previous attempts suffered from three major issues:

• Bandwidth
• Screen resolution
• User interface

Of these three, two are already solved: most broadband connections are now capable of streaming at least SD video, and increased resolution of HD TV makes web text quite readable. The third issue alone remains: the user interface.  Recognizing this, Hastings predicts a new generation of Nintendo Wii-like pointer/motion remotes to replace the primitive up/down/left/right arrows (and four dozen other never-used buttons) on today’s remotes (interestingly, Apple has recently filed for a patent on some technology for just such a device).

Look for this to be the big story at the CES show this January…

How much traction is online digital distribution gaining?  Look at what’s going on these days vis à vis DVD release windows…

Hancock Several months ago, Sony crossed a line when it announced it would be releasing the Sony Pictures/Will Smith vehicle “Hancock” over the internet (exclusively to the Sony Internet Video Link device) prior to the DVD release date of November 25.  A good move on Sony’s part: leveraging their unique position as one of the six major film studios and a major CE manufacturer to help drive sales of their new video over IP hardware.

Wall-E At the time, I wondered if there might be similar synergy (remember that word?) in an exclusive release of Pixar’s “Wall-E” exclusively to iTunes/Apple TV before the DVD release, given the close Disney-Apple relationship (when Disney acquired Pixar from Steve Jobs, he became Disney’s largest individual shareholder and was given a seat on the Disney board of directors).   As it happens, Disney/Pixar is planning to release “Wall-E” to Apple TV concurrently with the DVD release, on November 18th  (and will do the same for “The Incredible Hulk” on Ocotber 21st).

The Office On the TV front, last month the well-received Fox/NBC-Universal initiative Hulu announced the online availability of several NBC shows’ 2008 season premiers (including flagship comedy “The Office”) prior to their air dates.  I thought this was a particularly bold move, and was frankly surprised by the relatively little press attention it received (see my take on Virginia Heffernan’s theWB.com article here).

Iron Man Sony Pictures and Disney have obvious options for hardware partnerships (Sony and Apple, respectively), but what’s a Paramount to do?  Enter Dell: the two companies have just announced that for an extra $20, Dell customers can now have the Paramount summer hit (along with bonus footage) pre-loaded onto the hard drives of newly ordered Dell machines.  But while watching a sitcom on the computer is (marginally) acceptable, what about longer-form content such as movies?   Um, no.  Not gonna happen, at least not on any meaningful scale.

For that, you need two additional specialized pieces of hardware:      a television.     and a couch.

Dell Fact 1: Dell and Paramount know the limitations of movie-watching on a PC.   Fact 2: Dell also makes televisions.  It all adds evidence to the rumors of Dell rolling out even more CE hardware and (drumroll…) an iTunes-like service at some point in the future…. but to do that, Dell will have to do two things: retool its workplace-centric brand, and hire some new product design talent  (because I’m not sure Dell understands the benefits of simplicity).