Communications

What does China want?

Sun Bin has an excellent article on what the Chinese Communist Party (CCP) envisions for China in the medium term (20-50 years), based on his analysis of the messages in a recent Chinese TV documentary series, "The Rise of the Great Nations."

Many years ago, I read Paul Kennedy's The Rise and Fall of the Great Powers.  It appears someone in the Chinese leadership also read this book, prompting a CCP Politburo seminar on the subject in 2003, and the beginning of TV production project that took three years and resulted in the recent CCTV documentary.

The video (Chinese only) is now available on several websites (1-7, 8, 9, 10, 11-12) or bittorrent. (The latter server hosted in the US and run by exiled pro-democracy activist Hu Ping).  There's also some discussion (in English) here, but Sun Bin's article is the best analysis I've seen, including his summary of the key principles being pushed:

1. Great Nation = Innovation and contribution to its own people and the world
   ... In the Finale the head of Chinese Diplomatic Institute Wu Jianmin was quoted as saying something like this, "Great nation contributes to the world development, innovation is essential for making such contribution, and innovation is not possible if free thinking is constrained"
2. Aggression through force is to be avoided at all costs
   ... the program made it clear that the modern world order is one in which competition is by business and innovation, and contribution to humanity and scientific knowledges, not by military success.
3. Rule of law and building of a system
   ... in almost all episodes, especially that of UK and US, the rule of law were emphasized ...
4. Focus on internal development and building sustainable capabilities
5. For catching up 'great nations' (e.g Germany, Japan, Russia), the government must play the leading role
   This is perhaps the most controversal point in this program, IMHO, as it not only emphasized the role of government "planning", praising its contribution to the Russian industrialization in the 1920s-1930s, it also praises the Keysian economics which influenced the role fo the governement in Rosevelt administration -- my concern is that it is difficult and tempting to forget the fact that government role is needed only when the problem is so serious that the invisible hand is unable to cope with it, and Chinese bureacracts, like bureacrats elsewhere will soon forget the principle that government intervention is needed only as a last resort)
6. Together with the government role above, the views on USSR's "worthy experiment" on socialism and planned economy is positively ambiguous
   My speculation is that the Politburo could not agree on these issues and decided they will keep the conclusion open (again, this is assuming that the program is part of the state propaganda). Same can be said about when globalization will truly make aggressive war obsolete, as in the Finale it stated such time will come even though if not in the medium term.

If you are at all interested in China, I heartily recommend reading Sun Bin's complete post.

Broadband retrospective (humor)

This is The Future

Ah, if only...
From blaugh.com, comics of the blogosphere, where you can see other great panels like this.

How little YouTube pays for Internet connectivity

I just finished reading an interesting discussion of Web 2.0 issues entitled Web-Reloaded? – Driving Convergence in the Real World, by Arthur D. Little's Telecom, IT, Media & Electronics group in Germany (available free, in English, if you fill out this form which is in German).  But there were a few of their statements I questioned including the assertion that, with more than 80 million videos being viewed by day, YouTube's "bandwidth and service costs are estimated at between € 245-325K a month."

This is not the first time I've seen assumptions that YouTube must pay a lot for their Internet connectivity.  Last June (when YouTube was using 20 Gbps to support a mere 50 million videos per day, Keith McMahon suggested their ISP connectivity costs were at $512K per month.  Back in April, Forbes suggested that only 13 million videos per day might be costing YouTube $1 million per month.  Many others echo this idea of a large expense for "bandwidth." Even Wikipedia repeats the idea:

Some industry commentators have speculated that YouTube's running costs — specifically the bandwidth required — may be as high as US$1 million per-month, thereby fuelling criticisms that the company, like many internet start-ups, did not have a viably implemented business model.

I don't have inside information on YouTube's bandwidth costs but I do know something about peering thanks to many years of reading the Cook Report on Internet, e.g. this (subscription required), occasional tidbits from North American Network Operators Group (NANOG) meeting reports and articles like this by Bill Norton of Equinix.

I can assure you, as soon as YouTube's traffic became a measurable fraction of Internet traffic, probably early last spring, the tables turned and YouTube began to be able to negotiate free peering. In fact YouTube's peering coordinator Colin Corbett showed up at the June 2006 NANOG meeting to explain who they were and where an ISP could peer with YouTube.

The first peering candidates are second and third tier ISPs who have to pay Tier 1 ISPs for "IP Transit", i.e. for access to Internet addresses they can't reach directly or by peering with other ISPs.  As of June, YouTube was connected to the largest US public peering network, Equinix, in San Jose and Los Angeles and to the second largest, PAIX, in Palo Alto(?).  Any 2nd or 3rd Tier ISP with access to Equinix or PAIX would be investigating peering with YouTube in order to avoid paying a Tier 1 ISP for access to YouTube content.

As a service like YouTube grows to become a significant source of Internet content, their leverage with tier 1 ISPs increases. YouTube's traffic is very asymmetric – a very small request comes in and a very large video goes out.  If they are using one Tier 1 ISP for IP Transit service, other Tier 1 ISPs see this large asymmetry in their Tier 1 peering arrangements. Typically, Tier 1 ISPs interconnect with each other on a "free" basis as long as they have symmetric traffic.  Suddenly there's incentive for other Tier 1 ISPs to interconnect with a high volume asymmetric content source like YouTube.

The specifics of peering arrangements are murky as most deals under NDA, so there is no way to know for sure what deals YouTube has been able to negotiate.  But it's clear they've been at it for some time now and, even without the Google acquisition, YouTube's expense for Internet connectivity was not soaring but in fact was more and more under YouTube's control.

And of course, even if YouTube (now Google) had to pay retail for Internet connectivity, their potential revenue is so large, it would be a wildly profitable business.  But that's a different story.

Afghan update - Roshan signs millionth subscriber

With a third competitor, mobile phone adoption in Afghanistan is picking up.

Earlier this month, Roshan Telecom announced their millionth subscriber.  I took notice because I attended a wonderful presentation by their CEO, Karim Khoja, at CTIA last spring (and wrote about it here).

Some of his anecdotes were amazing.  To build a typical cell site they first have to clear land mines and then build an access road. Then they install redundant diesel generators and substantial fuel storage as there is no electric power grid. Today, Roshan's combined electrical generating capacity is the largest in Afghanistan.

Rural cell sites have armed guards to protect against theft of fuel. However, they've never had a cell site attacked - even when there is fighting nearby, all sides understand the benefit of communications.

Last April, there were two operators in Afghanistan, Roshan and Afghan Wireless Communications Company (AWCC) with a third license granted but not on-line yet.  In July, Areeba launched the third network and it appears they managed to sign nearly 100,000 subscribers by the end of the September. Blycroft Publishing puts AWCC's 3Q subscriber count at 476,000.

That suggests total Afghan mobile subscribers have grown from 1M to 1.475M in just six months.  That's still tiny in a country with a population of 31 million (CIA estimate July 2006), but it's the kind of jump you expect when additional competition comes into a market.

Alice & Bob - Cryptography humor

From Randall Munroe:

In a recent email exchange, Mike Monroe (Secretary & Treasurer of PICMG) pointed me to xkcd, a blog of wonderful cartoons on "romance, sarcasm, math, and language," that his physicist son has been publishing for some time now.  xkcd is famous enough to have it's own Wikipedia entry, so you may already have seen these.

I love the topics he covers.  As the disclaimer puts it

Warning: this comic occasionally contains strong language (which may be unsuitable for children), unusual humor (which may be unsuitable for adults), and advanced mathematics (which may be unsuitable for liberal-arts majors).

Mobile Advertising - It's very early days

In today's Mobile Marketing track at World Telemedia in Budapest, the first session was titled: "Making Money in a Free Content World: getting clever with ad-funding and beyond." This is an area just getting off the ground.  Purely anecdotal but interesting, was a video program that flipped from paid downloads to ad-funding, i.e. free to the consumer.  In the 12 months prior to flipping, they had 5000 downloads.  In the 12 months after, they had 24 million downloads.

But slow speeds and limited browser context (as mentioned in my last post), severely limit the ability to have click through ads. Most of the mobile ads discussed this morning were banners using publisher pricing like a magazine, i.e. cost per thousand (CPM) views, rather than the advertiser-friendly Internet approach of cost per click (CPC). This is very lame and unlikely to foster the phenomenal growth rates we've seen for Internet advertising.

While I don't have EU figures at hand, I thought these figures from a recent Morgan Stanley report on US Internet advertising were revealing,

	Unique Users (000s)	Total Pages Viewed (MM)	Total Minutes (MM)	US Ad Rev/ User (2Q06 annualized)
Total Internet	173,428	462,186	310,466	$31.44
Google	107,365	10,264	5,232	$12.28
Yahoo!	129,724	39,878	42,715	$8.65
Microsoft	119,376	18,721	22,125	$3.25
Time Warner	120,276	18,000	36,106	$1.62
MySpace	55,849	34,988	11,503	$1.32
eBay	78,991	10,972	6,480	$0.33
YouTube	20,759	1,386	1,047	<$1.00

especially when compared to US mobile ad revenues (mostly SMS marketing) of well under $1 per user.

Chetan Sharma has a good overview of mobile advertising status and prospects, here.

Mobile Marketing at World Telemedia

I was in the Mobile Marketing and Branding track at World Telemedia for most of today, including speaking in the last two sessions of the day.  I hadn't thought about until today, but over the past year I've become somewhat connected in this field – I kept running into people I know and have just seen at other recent conferences.

Some forms of mobile marketing are highly developed.  I refer of course to SMS and WAP push-based systems which have evolved from simple SMS broadcast to highly targeted campaigns, based on contextual and behavioral information. It's just economics – even at bulk rates, SMS costs money and everyone is measuring their ROI.  Also, on the WAP side, all the web analytics that can be applied are in use, so rich click streams are routinely captured.

But there are significant limitations in the mobile space compared to the web. About the only things common across all handsets are voice, SMS and, hopefully, a WAP browser. But that's a browser with only one context.  If you have clickable advertising in your handset application, and the user clicks, they are gone from your application and must restart and possibly re-login upon return. Despite problems, there many companies pressing the limits.

The situation with the mobile operators is less advanced – at least compared with data-driven marketing leaders like Walmart, Tesco or Amazon.  Worse, the big issues for operators are acquisition costs, ARPU and churn.  Despite a public desire to drive data and content revenue, operators can't afford to jeopardize their cash cow, which is voice and person-to-person SMS. So they are extremely cautious.

What was my topic?  My title was Leveraging Telecom-Specific Info to Drive Consumer Behavior.  By telecom-specific info, I refer to Call Detail Records (CDRs) – a rich set of information phone companies have been capturing for 100+ years, but have never used for anything besides billing.  Unfortunately, I'm not prepared to write up the details right now.  Sorry, you had to be there...

World Telemedia - Mobile TV

The second session of the morning was Mobile TV & Video: Moving pictures.  It’s interesting to see the EU perspective, as I’ve just come from Hong Kong where at least one high quality, ubiquitously accessible mobile TV service – based on 3G video calling – has been up and running at Hong Kong CSL for many months.  Asia also hosts a variety of interactive video response services based on 3G video calling.

Key take away for Mobile TV in Europe – it’s very early days.  Indeed, it may be 2012 before we see widespread use of mobile TV, before the spectrum needed for video broadcast, e.g.DVB-H, DVB-T, etc., has been cleared and efficient broadcast to mobiles is working.  Multiple speakers pointed out the problems with today’s services based on video over packet services – QoS problems, outrageous packet data charges, codec/handset incompatibilities, DRM issues.

The breakthrough solution being promoted by Jeremy Flynn of D2See and by Mark Fitzgerald of MXTelecom, is 3G video calling.  Of course, this is what’s already running in Asia, supporting my view that hot new mobile applications almost always start in Asia, emerge in the EU 12–18 months later and then make their way to the US eventually.

It’s true, 3G video calling has been a flop as a person-to-person video calling service.  When it is used at all, it’s used for see-what-I-see purposes, i.e. two way audio and one-way video – “Look Grandma, the baby has learned to walk.”

But used as a means of delivering video content or receiving user created video content, 3G video calling is ideal.  You can make the connection in seconds.  Once connected, the video quality is constant (as it’s running over a 64 kbps circuit-switched data connection).  It’s inherently streamed without local storage, so no DRM issues.  And all 3G handsets have an inter-operable 3G video calling client built in.

 

World Telemedia Budapest - first impressions

Evolving approaches, enduring themes.  The opening key note is about interactive broadcast TV, i.e. using your mobile phone to interact with, and download content from TV programs like Big Brother and Deal or No Deal. 

This is my first time at a World Telemedia conference.  It’s billed as Eastern Europe’s number one D2C (direct to consumer) conference, but the organizers are UK-based and a majority of the speakers and companies are from the UK and other parts of western Europe.  Many names I’m already familiar with.

Walking through the exhibit hall to get to the opening session, I see booths touting mobile content interactivity via SMS, MMS, IVR, WAP and video calling; WAP chat; global premium numbers; global premium SMS; universal 3G video calling platform; and so on.

What strikes me the most is the parallel with my first Audiotext show, in Las Vegas in 1990!  Yes, the technology has diversified enormously, but the underlying themes are the same:  adult content, chat and dating, celebrities, sports, games…, and the pervasive problem for independent content providers is billing, thus the focus on premium numbers, short codes, and alternate billing arrangements.

 

Connect 2006 Group Photo

Photos - Day 1 at Connect 2006 Kuala Lumpur

Here's a set of photos that Jan Geirnaert took, including many of the slides shown.

As a byproduct of the way I loaded these into Flickr, you need to start at the bottom and work your way up if you want to understand the sequence.