Tuesday, October 03, 2006

The IPTV conundrum in Asia

This blog is a reaction to a White Paper I read called “The IPTV conundrum in Asia” by Parks Associates.

Hong Kong has 500,000 IPTV subscribers. Part of this may be due to the density – it is much easier and cheaper to wire inside a building than to a building. Once you are inside, you can light up 20 floors times four flats – 80 subscribers with 1000 feet of cable. Try that with 80 homes in say, Milwaukee, Wisconsin and you are talking about covering at least 10 acres and 10 miles of cable.

Of course the main reason that many Asian countries including parts of China such as Hong Kong, but most notably Korea are much further along than the US and Europe in Broadband to the home. A notable exception – India.

Another good reason is that Asian content providers, being more realistic than their western counterparts have been willing to give their content to the IPTV providers. Again, a notable exception has been Indian content-owners but that has changed recently.

According to the paper there are two rules of thumb for Asia (I believe this applies even in the US):
1. Consumers may like the instant gratification offered by video-on-demand, but they won’t live without the linear (i.e. traditional) format of TV today.

2. “Cheap is good: free is better” – consumers are hesitant to purchase expensive equipment from operators. Low-cost set-top boxes have thus become a key ingredient for success.

I agree completely with the first item – it is a well known fact that consumer behavior takes as long as a generation to change. The second point is also well-made, but the conclusion is arguable. Instead of low-cost set-top boxes, the providers need to combine set-top boxes with PVRs or at least DVD players. And these set-top boxes need to be operator-neutral, or customers will balk at paying to be tied to one operator.

An interesting finding of the paper was that “Asian consumers are more comfortable than other people in the world ….(with PC-TV convergence, which) … has already occurred in terms of how consumers view the various platforms. I wonder if this has to do with the fact that all these gadgets appeared in homes at about the same time unlike in the west. It could also be that Asians, unable to resist a bargain (it is in the genes), see potential savings in the convergence. As my uncle used to joke “Indians will fly to the US (or Europe) standing if airlines offered it inexpensively enough”.

Another very interesting point – the appeal of an interactive program guide is actually lower in Asia than the US or Europe. I have no idea why this should be so, perhaps a reader can enlighten me.

One seemingly unimportant sentence in the paper stood out for me – “…getting Asians to pay for the features they desire will be difficult”. Wrong! However, getting them to pay for features that the providers want them to want (because it fits their business model) will be. At least from the Indian experience, it is clear that if there is compelling value, and people want a feature, they are willing to pay for it. Take for example DVD players. I have spoken to several industry experts and front-line businesspeople who said the rush to buy DVD players was unprecedented. In just a few years DVD players have become more pervasive than VCRs ever were.

And that is the bottom line. Give consumers what they want – and they will pay for it. What a concept, huh?

So what do consumers want? They want somebody to manage all the content that is available to them. In addition, they want niche content – programs in their language (remember India has 22 languages) even when they live outside the normal reach of that language.

It also means being able to watch anything anytime. The intelligent functions of a PVR will definitely sell well. In our demos, people love the idea of being able to record all episodes of their favorite shows, or simply pausing live TV.

There is a growing segment of Indians who are traveling both for business and pleasure, and many that have returned from long foreign sojourns who want more than what is available on Cable TV. This is what India’s marketer call the “Aspirational Segment”. They want not just Hollywood movies, but the content that relates to the lifestyles they left behind in the US or Australia or elsewhere.

More on that later…

Learn from TiVo

This is an old one that I had published in 2005 elsewhere - but it is still relevant.
One thing we can learn from TiVo’s stumbles is to not ignore the customer. And more importantly, to know who your customer is.

TiVo avoided the consumer – they still do. They did not have a good process to support sales. And their support is abysmal. Did you know they removed email support? The on support available is by phone – and it is only an automated system – rarely do you reach a person. And of course if you bought a TiVo branded by another company you are completely out of luck. Here is a company that seems to be going backward not just technologically but also in customer service attitudes.

As I have mentioned, they knew that people who buy TiVos become TiVo evangelists, yet they did not mine these people. Check out TivoCommunity.com – the tone there is more like a Linux mailing list than a user community. Any user who asks a “newbie” question gets jumped on.

I myself have sold people on the benefits of a TiVo (grudgingly of course – who wants to make the competition stronger), and they have gone out and bought them. Not only that they call me up and tell me how they love their TiVo – these are rational people, mind you! I evangelized TiVo for a while before I realized that the business I wanted to be in needed something like a TiVo to use as a platform. And I decided to build my own.

TiVo should have set up a selling process that demonstrates the benefits of TV. Try asking someone at Best Buy or Circuit City about a TiVo, and the most you will learn is that it’s like a VCR with a hard disk, and that you have to pay $12.95 a month for the service. Demonstrate it once, and you would have a stampede that your cashiers could not handle. Selling anything with many and complex benefits requires salespeople, not shelf stockers.

But one really important things that many surveys and industry experts have pointed out is this – TiVo’s best customer is not he 18-35 male electronics and gadgets buyer. Surprised? I was too. Although the buyer is usually a male, most users and evangelists are their female counterparts, and often have children. My wife, who is as close to a Luddite as anyone I have met, loves TiVo. It’s simple – she can control how much TV the kids watch, what they watch, and most important, when they watch it (on weekends after homework is done). And does TiVo market to this segment? Not a chance. In their place, I would have put together a Tupperware-type network to sell TiVo – one mother to another. Anyway, no one has offered me the VP Marketing position there, so I will make my own!

I think early on TiVo decided their customers were going to be the DirecTVs of the world. That is why they kept courting the satellite and cable companies. But when push came to shove, they decided they were too good to partner with the likes of Comcast. That really triggered the resignations you heard about after CES 2005. Suddenly they are mouthing off about how TiVo is more than just a set-top box, and consumers want more than the ability to store programs and fast-forward and rewind. The CEO Mike Ramsay went so far as to say that the Cable TV business was not the business of the future. Well Duh! And it took you how long to figure that out? TiVo should have figured that out long ago – many people including yours truly did. And they were in that business! That’s like finding no WMD in Iraq. Again, I can’t resist it – Duh!

Of course, it’s not all over for TiVo. Ramsay is right about disruptive technologies like TiVo and even better technologies like L3 destroying the Cable business over time. But mind you, we are talking twenty or more years for the cable business to go away. And then you will have Internet TV businesses that… oh, yes, might be called Comcast and Time Warner…

Tuesday, August 08, 2006

Current & Future IPTV Models

Most people in the video content and delivery industry believe that the method of delivering content in the future is going to shift primarily towards the Internet. However, it is important to create a roadmap for acceptance in the industry – consumers, manufacturers, content owners and advertisers must clearly see a benefit in the short term so that they will adopt the new technologies in the long term.

In the current IPTV (Internet Protocol TV) model, video is generally stored on a server and then “streamed” over an Internet connection to the user when requested. Hence IPTV can simulate “channels” similar to Cable TV, but it can also provide Video-on-Demand (VOD). In fact, IPTV has an edge over Cable TV in VOD applications. An IPTV customer is either given or sold a set-top box by the ISP (Internet Service Provider); this box being used as a “tuner” that interacts with the service and displays video on TV. In the most sophisticated IPTV applications, the box is also used as a payment gateway.

Current IPTV models – both business models and technical models, are essentially similar to Cable TV models with some minor variations. While the current models are workable solutions, there are several missed opportunities with both the technology and the business models in use today.

First, Streaming Video, although designed for the Internet, is unreliable in practice. If the video is not on the ISP’s network, its delivery to the consumer cannot be guaranteed. Sometimes, even when the video is on the same network as the consumer, there can be quality issues when concurrent usage is high. Also, when the content is outside the ISPs network there is a significant additional cost to the ISP in the form of peering and bandwidth charges. Thus many IPTV providers are forced to negotiate rights to content in order to place the content on their own network – this creates a significant and ongoing operational burden as well as substantial upfront costs.

Second, current IPTV models generally do not support features that are intrinsic to an Internet based delivery system. An excellent example is file sharing (peer-to-peer or P2P), which can result in a significant cost saving in bandwidth if managed correctly. It also provides a strong marketing opportunity due to its viral nature. However, no current IPTV system supports file sharing because the technical problems of legal file sharing have not been solved until now (L3 technologies do solve the problem).

Third, IPTV models that deliver streaming video using the “Client-Server” approach require significant upfront costs to launch successfully. The primary reason is the cost and operational issues of deploying servers near the users. This also impacts the ability of IPTV providers to quickly expand into new geographic markets. An IPTV provider needs to be able to enter this market at a very low capital cost; else this cost can become a barrier to success.

Fourth, using streaming servers predicates insertion of ads before the content reaches the user. This prevents extremely granular targeted advertising. Cable TV is capable of ads targeted at the neighborhood level. However, if correctly deployed, IPTV can target advertising at a significantly higher level of granularity than Cable TV. For example, advertising can be related to (a) the content being viewed (b) the location and preferences of the user, and (c) the time at which the program is being viewed (as opposed to when it is broadcast). This is very similar to the kind of advertising model available on web pages, and it will not be long before video advertisers expect or even demand it.

Finally, getting consumers to accept IPTV initially is a significant marketing challenge – they are generally satisfied with their Cable TV or Satellite TV (DTH) services. Although IPTV has significant advantages over Cable TV in the long term, any IPTV solution that hopes to introduce “yet another box” into the living room, must have a compelling current selling proposition to a consumer. Also, a “go-it-alone” approach in IPTV means that the IPTV provider has to negotiate a critical mass of content rights. As described above, this can be operationally burdensome, expensive (since most content owners are unconvinced of IPTVs potential, they often demand a premium for those rights) and a barrier to quick entry. Hence, an ideal solution is one that coexists with the current paradigm/providers of Cable TV, but introduces future distribution models in the form of IPTV.

Thus, what is needed is a technology that propels IPTV to a higher level than Cable TV can achieve through their model; a compelling set of current features that are compatible with the way people watch TV today; and the promise of new features that will improve their TV watching in the long run. At the same time, the technology must be so attractive to both content and advertising that these industries will “pay-to-play” rather than the other way round.

Monday, May 01, 2006

Why Streaming Video is wrong for IPTV

In a previous blog I had explained the basics of IPTV. In this, I want to explore one method of deploying IPTV and why it is doomed to failure. Let me also point out that Streaming Video and Streaming Audio are parallel models and together are often referred to as Streaming Media.

Lets start by noting some of the well-known methods of transmitting video across the Internet. The simplest and oldest method uses the File Transfer Protocol (commonly called FTP). A recent and very popular method called BitTorrent allows very fast transfer of files under some conditions – for example, it excels in peer-to-peer networks.

Now lets look at Streaming Video – again, this bit is a rehash from the previous blog, but it is important, so please stay with me. Streaming Video is a method of transmitting video files so that a portion of any video starts to play in your video player (for example Windows Media Player or Real Player) while the rest of the file is being received. All other commonly used methods require that the entire file be downloaded before it can be played.

Streaming Video essentially creates a “buffer” which fills up as the video is received. At any time, if the connection becomes slow or is briefly interrupted, the video continues to play uninterrupted from the buffer. Meanwhile, in the background, the player attempts to download more of the video into the buffer.

As I have said before, that is the theory, but in practice, interruptions and slowdowns of Internet connections are rarely brief. So, there is not enough of a buffer to supply a continuous stream while the video is being viewed. The video becomes choppy and grainy, sometimes freezing altogether - resulting in a poor viewer experience.

Let us say you are lucky enough or rich enough to buy a super-fast connection to the Internet. All you need to worry about now is that people in your neighborhood are not using their Internet connectivity too much when you want to watch a movie on IPTV.

“Ay, there’s the rub!” as Hamlet said. However, your ISP is not conflicted, their entire business model is based on shared bandwidth. In fact, Internet connectivity is always shared at some level or the other. Most commonly, connectivity is shared at the neighborhood level. And typically, most people in a neighborhood will go online or download movies at more or less the same time of day. Thus, at the very time when people want the best connectivity, the system will be under maximum load.

Internet connectivity is also unpredictable. In some ways, it is like the weather – there are just too many variables. A small storm in one area can result in massive upheavals thousands of miles away. When a stream of video travels from a providers server to a users PC, it could travel through a dozen networks and multiple nodes within each network. And the number of networks may or may not be related to distance. For example, even if the server and the users PC are a few hundred yards apart, the video might actually travel through half a dozen networks. Each of these networks could be having a massive upsurge of traffic just when your movie reaches a cliffhanger moment – literally.

It is for these reasons (and a few others) that the promise of delivering video-on-demand through the Internet has been largely unfulfilled. Despite improvements in Internet connectivity and speed, the shared pipes are still not big enough to reliably deliver a consistent and reliable stream.

Despite all this, most IPTV providers still use Streaming Video. Why do they want the worst possible protocol for transmitting over an unreliable and narrow pipe? For the single reason that, because it starts to deliver video to the user immediately, most people think that it is still the best way to achieve Video-on-demand. Other known protocols will force the user to wait until the entire file is downloaded.

But are there alternatives? Yes, and it is obvious once you think about it. The best way to deliver Video-on-demand to a consumer is to have the video already in their home, locally.

There is of course yet another way that Streaming Video can work. That too might be obvious after reading the above. If the Streaming Video provider is also the owner of the network, that provider then has far more control over the delivery of video. Let us say AT&T is your ISP, and they also provide content to you in the form of movies. Then, AT&T can design the network so that the movie is on a server near you and also ensure that there is sufficient bandwidth between the server and you. Expensive, but it can be done by running fiber to you door.

Unfortunately, such a model has many other issues including the inability of ISPs to obtain content easily. Most importantly, this model requires owning fiber (or at least copper) to the customer. And that limits any company from establishing itself in multiple geographies quickly. It limits them to serving specific geographies in an increasingly mobile world of fragmented audiences. More on this in another blog.

Yours truly is also working on a method that will simulate the benefits of Streaming Video but will rely only partly on the speed or reliability of the connection. Wait and see.

Copyright (c) Anil Gupte 2006, All rights reserved.

Sunday, April 30, 2006

What is IPTV?

So many people have asked me what IPTV is about, that I have finally decide to write it down. Read on...

In IPTV, IP stands for Internet Protocol. In the world of networking, this protocol is the third layer out of seven layers that together create a complete network capable of transporting content from point A to point B on the Internet. This content can be text, pictures, voice, and, this is where our interest lies - video as well.

Although most people think understanding IPTV is complicated, the basics are very simple. At one level, all we are talking about is the transmission of a file from a server on the Internet (the same server that might also host and transport web pages) to a users PC. In fact it is not very different from the process of serving up web pages - with one exception.

In other words, and this is important, the transport mechanisms of the Internet do not care what you are transmitting. So, a video file can be transmitted in exactly the same manner as say, a web page (an HTML file). The difference of course is that a video file is going to be much, much larger than just about any web page, hence taking (you guessed it) much, much longer to move from point A to point B on the Internet (the exception I was talking about).

You know that famous phrase about a picture being worth a thousand words? True or not, a picture can take about a thousand times more storage than a word. Now, multiply that by 30 (the NTSC frame rate, 25 for PAL – if you are not sure, never mind approximations are OK) to get the amount of information in one second of video. Impressed? Now you can understand why everyone thinks transmitting video over the Internet is a “big” deal! It has less to do with complicated technology than size (i.e. the amount of information transmitted). Of course video files are compressible depending on the type of video in the file, but the point is that they are many orders of magnitude larger than almost any other file.

About three paragraphs ago, I said, “…a video file can be transmitted in exactly the same manner as a web page…” Luckily for me, I did not say “is transmitted”, and that gives me the opportunity to talk briefly about “Streaming Video”, which is often (incorrectly) used interchangeably with IPTV.

First, let me point out that there are many ways in which a video file can be transmitted over the Internet. And most methods of file transfer can be used with most files. The simplest and oldest method of transmitting files uses the File Transfer Protocol (commonly called FTP). A recent and very popular method called BitTorrent allows very fast transfer of files under some conditions – for example, it excels in peer-to-peer networks.
Streaming Video is a more specialized method of transmitting video files. In this method, a portion of any video starts to play in your video player (for example Windows Media Player or Real Player) while the rest of the file is being received. To do this the software essentially creates a “buffer” which fills up as the video is received. At any time, if the connection becomes slow or is briefly interrupted, the video continues to play uninterrupted from the buffer. Meanwhile, in the background, the player attempts to download more of the video into the buffer.

That is the theory anyway. In practice, interruptions and slowdowns of Internet connections are rarely brief. That is why Streaming Video cannot succeed unless, and this is important, unless the people doing it also own the network.

And you cannot own the network unless you think small. But more on this in a later article or blog. Stay tuned.

Copyright (c) Anil Gupte 2005, All rights reserved.

Sunday, April 09, 2006

Bollywood's "Long Tail"

Forget squeezing millions from a few megahits at the top of the charts. The future of entertainment is in the millions of niche markets at the shallow end of the bitstream.
The Long Tail

To understand what I am about to say, it may help to know about The Long Tail, first an article in Wired, then a blog, then a best-seelling book. The author Chris Anderson talks about “The Long Tail” - in content, it refers to the niche audiences for certain types of content i.e. "hard to find". And there is no doubt that there is a long tail for content - having proved its existence at iCinema.com, for example.

At iCinema.com the initial audience served is the approximately 3.2 million Diaspora of Indian origin in North America (plus similar numbers in UK, Australia, Africa and the Middle East). Most people can immediately grasp that immigrant Indians in the US want content from Bollywood. After all, if Bollywood has created a buzz even among Middle-Americans, it must be the what this Indian Diaspora is looking for, right? Wrong! It goes deeper than that. Yes, of course there is a demand for Bollywood content, and it cuts across the entire group. But, there is a thriving movie industry in every one of the seventeen regional languages of India (no we are not talking dialects – there are some 300 plus of those). So there is an even longer tail beyond Bollywood audiences, and if you do some “area-under-the-curve” math, you will find those numbers are even bigger than the numbers for Bollywood.

L3 Media is all about “The Long Tail”, and more importantly about monetizing it.

Okay, so you have an audience for Marathi (substitute your favorite language here – Swahili, French, Russian) films in the US. The average market size for each of the Indian languages is roughly 200,000 (about 50,000 households). How do you deliver content to this audience; and more important, how do you monetize it?

But first, let us talk a little bit about how you cannot. Take any one-to-many, traditional, system such as Cable TV or Satellite TV. There simply are not that many channels that these technologies can deliver. If you wanted to service all the Indian languages (or all the Eastern European languages, or all the African languages), you would need about twenty channels even with a single channel per language. And one channel for all the movies, news, interviews etc. that this group might want is simply not enough. There are 65 new movies made in Marathi each year, and the numbers are higher for some of the other Indian languages.

Even if you could shovel all the content for each language into one channel there is only so much of it that could be made available at prime-time. You could hope that people would record it on their PVRs, but then how would you monetize it? Certainly not from commercials, because once the content is on a PVR, there is no mercy from that thumb on the fast forward button.

Coming back to those 50,000 households to whom you want to deliver content. If you leave out the one-third or so who live in New Jersey or Los Angeles or one of the major metro concentrations of Indians (just look for the “paan” stains on the walls) you have approximately 200 households per ‘B-size’ city (another third) and then a dispersed remainder. Tell me Time Warner or Comcast cares about this market, and I’ll sell you a bridge in Bombay. But aggregate them online and suddenly you have a market! Even at a conservative $10/month average per household (alternative entertainment), you have a $6 Million market. But what if you could give them tons of free content (supported by ads of course), you could see a business that is 5-10 times that. Only problem is, with the current paradigm it is practically impossible to do the latter. How do you insert an ad for the sale at Boston Store in downtown Milwaukee into the current episode of “Kyonki Saas Bhi Kabhi Bahu Thi”. (If you just went “Huh?”, think “Dallas” or “As the World Turns” 20 years ago). The answer to that question is – you can … with L3 technology.

Copyright (c) Anil Gupte 2006, All rights reserved.