Jeremy Allaire’s Big Picture

From ContentNext:

We’ve seen a significant shift in the past year in the number of mainstream publishers offering pay and ad-supported consumer video services. Browser and PC based platforms are getting richer and richer in terms of the kinds of user experience that a publisher can achieve, and consumers are starting to open their wallets. Simultaneously, the emerging connected home world is getting more interesting, and the next few months are going to see an enormous string of new product introductions and announcements centered on using both personal and commercial media (audio and video) from the Internet and your PC and consuming it in the living room. This will get real mainstream attention, not just techies and insiders.

Through all this, though, I’ve tried to explore bigger issues around how video will evolve as a content type on the Internet. Most companies are narrowly focused on what will happen with Big Media in its transformation to digital distribution (yes, this is very important and a huge opportunity), but a more interesting question is what kinds of new video content businesses will emerge or rather thrive in the Internet distribution age, that perhaps could not in traditional retail video distribution (e.g. cable and satellite). Again, a lot of the “old models” for thinking about this — e.g. aggregation; bundled subscriptions; carriage — seem inappropriate in the Internet world, which is deeply decentralized, and built from the bottom up on hundreds of thousands of communities of interest, reflected in content and online services. There’s a lot to learn from search engines, xml feeds, weblogs, etc. that can apply to how video will be distributed and monetized online.

Now that video can be produced cheaply and with reasonable production values, and now that it can be affordably distributed and perhaps even easily monetized, will we see an emerging new class of “video site producers” rather than classic textual content. In 1994 when the Web really emerged, it helped bring forth an explosion in the amount and richness of text that was produced and available globally. I believe we’re at the front-end of a very similar curve in video, and this world / opportunity is not going to look very much like how we as consumers find, acquire and view video today.

Jeremy adds, in a subsequent post:

I absolutely expect to see an emerging class of video publishers and studios, including spin-outs and startups that come out of the existing production value chain in mainstream media. In general there will be a disaggregation of distribution — e.g. many significant brands and productions that are “carried” by cable networks, who are in turn “carried” by cable operators, will have an opportunity to directly distribute their products in more innovative ways, and potentially more profitable for themselves. This will range from individual producers to significant new production businesses, as well as existing established publishers, programmers and brands.

There are significant challenges in helping consumers find, review and select video media. Certainly nothing as powerful and deep as Internet-wide text search exists in this field. Even in the most advanced arenas for aggregated video (e.g. cable operators with digital cable systems and VOD), there are virtually no good tools. This will be a limiting factor in how easily the medium meets consumer need.

Cost issues are being addressed adequately, and the price/performance curves on both bandwidth and storage (in the network and the home) suggest an attractive basis for economical video distribution.

Cheaper PCs in India

Professor Sadagopan writes:

In the Mobility Event organized by CNBC & Intel in August 2004 at Mumbai, I had made an observation When Maruti 800 can be financed at Rs 2,500 per month why not a PC at Rs 250 per month? Since CNBC broadcast the program many times on TV some of my friends watched it. In fact Dr MM Pant from Delhi even informed me that HCL has an offer for Rs 799 per month PC buying option. Today I am pleasantly surprised that PCS is offering a student PC at Rs 499 per month. More interesting is the fact that Indian Bank is financing it. Of late Public Sector Banks have been keeping a long distance from doing any financing of IT products and services. This launch is particularly interesting, as it will get many students from the poorer sections of the society into PC reach. If only BSNL can offer Rs 99 per month wireless access for 3 hours per day it will be further push towards computer literacy for poor students.

From an earlier post:

Maruti 800 is the largest selling people car with an installed base of 2 Million. Recently Maruti announced a financing scheme Just Rs 2,499 per month. It brings Maruti 800 car to the reach of the common man. Maruti 800 price is around Rs 200,000. PC prices have plummeted to Rs 20,000 or less. Why not Banks or even hardware vendors (through their Finance companies) finance PC purchase by individuals at Rs 249 per month. Our PC sales just crossed 3 Million while China consumes 8 Million PCs per year. Such a scheme will bring computerization benefit to small businesses. Any one listening?

Perfect Search

John Battelle is writing the last chapter of his forthcoming book and is brainstorming some ideas:

Imagine the ability to ask any question and get not just an accurate answer, but your perfect answer an answer that suits the context and intent of your question, an answer that is informed by who you are and why you might be asking. The engine providing this answer is capable of incorporating all the worlds knowledge to the task at hand be it captured in text, video, or audio. Its capable of discerning between straightforward requests who was the third president of the United States? and more nuanced ones under what circumstances did the third president of the United States foreswear his views on slavery?

This perfect search also has perfect recall it knows what youve seen, and can discern between a journey of discovery where you want to find something new and recovery where you want to find something youve seen before.

Also check the comments on John’s post.

Huawei’s Secret

Ninad Mehta points to an article in Light Reading:

The company is well on its way towards that goal, with $2.2 billion in orders during the first half of 2004, according to RBC Capital Markets analyst Mark Sue, who visited the company’s headquarters recently.

“Evident by its well-organized tours offered to customers and media, Huawei is making a concerted effort to improve its image,” Sue wrote in a note to clients on Monday.

Behind the slick tours, though, the secret to Huawei’s success is cheap labor — a benefit that eventually rises to the top line. “Huawei will often discount its products 15 percent to 30 percent vs. its competitors,” Sue writes, “but due to its cost structure, Huawei may have similar gross margins.”

While 12 percent of Huawei’s revenues go to R&D, some 46 percent of Huawei’s 22,000 employees are focused there, Sue reports. “Whereas a Western vendor may be able to affordably send one engineering representative to a customer, Huawei can send 4 or 5 at a comparable cost.”

Will Huawei’s R&D manpower matter outside of China? You bet. Huawei says 80 percent of its massive R&D workforce is deployed in software development — a crucial piece of the rapid service provisioning and opex savings that carriers crave.

“Huawei’s R&D costs can be as low as 30 percent that of competitors,” Sue writes. “Consequently, Huawei may continue to price aggressively, especially as it attempts to build initial footprints at large carriers.”

Next-Generation Infoware

Jon Udell writes:

I’ve been using Flickr, and writing about it, for the same reasons. Likewise Among other virtues, both exhibit a really important one I haven’t mentioned yet: you own your data.

When I think about meshing my own data with an infoware-style service, there are two key strategies I need to consider:

1. The entry strategy. In the case of, it was easy to weave my own stuff into the service. Using the procedure I detailed here, tags that I maintain on my blog entries are automatically sprayed into With little effort, I was able to create hundreds of integration points between two complex information surfaces — my blog and This was so effective that I decided to use for tag surfing of my blog.

2. The exit strategy. With first-generation infoware services it’s hard or maybe impossible to retract the information you’ve given them. Second-generation infoware challenges that notion. You can’t delete reviews you write for Amazon, which is why I’ve never written one there. (Instead I write about books on my own blog where services such as All Consuming can find them.) But you can delete links you submit to or photos you upload to Flickr.

From the user’s perspective, and Flickr support near-optimal entry and exit strategies. You can deeply and automatically mesh your own information with them. And you can undo that meshing. Participation in the services is thus an “at will” arrangement. If you maintain well-structured information, you can as easily mesh it with another comparably-equipped service. So the switching cost, as economists say, is low.

Microsoft’s Future

Steve Gillmor writes after watching a Jim Allchin video:

There is somebody to whom Gates and company can compare themselves– an updated version of Microsoft, a hungry team of up-and-comers and alumni hitching a free ride on their browser platform, unencumbered by a business model that limits its platform to a subset of the network and innovation around RSS and other disruptive technologies.

In a world of Longhorn Interruptus, a free, advertising-subsidized, combined e-mail and RSS aggregator client will have an even better chance of eating away at Microsoft’s control of the desktop. Gates inadvertently acknowledged as much when he summed up the new plan: “So now, we’re doing the search stuff in Longhorn ’06, and then we’re releasing WinFS off-cycle as a development platform and as sort of an information management shell synchronized with a release of the database server.”

Meanwhile, Sun’s Java Desktop System, Mozilla Firefox, Apple’s SafariRSS, and even Chandler, Mitch Kapor’s long-awaited open source collaboration system, will interconnect with Alchemy, peer-to-peer VoIP messaging services such as Skype, and componentized enterprise apps delivered as services by an alliance of software-as-a-service vendors.

In other words, Microsoft will spend 2005 and 2006 competing against itself with what will likely be a free downloadable (a la XP SP2) pruned version of Avalon and Indigo while delaying a competitive response to the Google-led hosted alternative to the Office “information management shell” to 2007 and potentially beyond. No wonder Allchin asks: “Does that make sense?” If Longhorn, the latest incarnation of Allchin’s 1995 Cairo unification of Windows and SQL Server, slips again, he may not be around for the answer.

TECH TALK: An Entrepreneur’s Growth Challenge: Roadmap

When I started thinking about what I wanted to do next in 2001, I had two building blocks: a small 15-person Linux-based messaging solutions business in Mumbai, and capital. That was not good enough. I am one of those people who need a reasonably clear vision of tomorrow and what we need to do before I can start. So, even as the messaging business has grown incrementally, I have been trying to envision the future.

I started with the premise that I will create solutions for emerging markets. Given by base in India, there was no way I could build products for the US or other developed markets. Those markets were far removed from the reality that I saw around me everyday. And I was not prepared to shift base to the US to be closer to the markets. Also, I was very clear from the beginning that I wanted to create a product-centric company. Outsourced software and services did not fascinate me. (Add to the list of starry-eyed entrepreneurs endlessly fascinated by the business models and success stories of the likes of Intel and Microsoft.)

The first idea in 2002 was to create a low-cost eBusiness suite for small- and medium-sized enterprises available for something on the order of Rs 250 ($5) per user per month. As I thought through this idea, I realised that we would have a tough time selling this in India there just weren’t enough computers in SMEs. The low installed base of PCs meant that any ASP-type solution would have a slow offtake. This problem also pointed to an opportunity the need to create low-cost computing solutions to counter the problems of high hardware and software prices. Between the two, the desktop cost was the first barrier.

Over time, as we thought through the problem, a possible solution came to the fore server-centric computing built around thin clients, thick servers, open-source software and remote server management. So, we created Emergic Freedom a Linux-based server-centric computing platform. But we barely made a dent in the market. I realised that people and companies had their own solutions: for hardware, it was non-consumption, while for software, it was piracy. If we were to succeed, we would need to take a look at the complete ecosystem of hardware and software.

This is where emerged the idea of the Rs 5,000 computer. My initial approach was to look at using old and refurbished computers, but these were hard to find in India. A stiff import duty made this idea a non-starter. If we had to solve the hardware problem, there was only one approach: the 5KPC would have to be built! That is one of the paths we embarked on a few months ago, even as we started thinking about ideas like grid and utility computing which would benefit from the ever-improving communications networks.

This slow, incremental evolution of ideas has now been encapsulated into a nice elevator pitch for Emergic: CommPuting at the price of a Cellphone hardware, software, broadband connectivity, content and support for Rs 700 per month. The focus is on the next users in the middle of the pyramid in emerging markets with India as the starting point primarily, homes, SMEs, educational institutions. The vision is not very different from Microsoft’s in its early days a computer accessible to every family, employee and student. My belief is that the time has come to reinvent computing for the world’s next billion users. The vision is now there, and there is a roadmap to take us from here to the future. It means building the thin clients, developing the backend grid computing platform, and creating the services which can make the solution desirable for the next users. It is now time to move ahead.

Tomorrow: People