Ruby on Rails at SDForum Tonight

Tom Hill, one of our senior engineers at Zvents, will be speaking about Ruby on Rails tonight at the SDForum Emerging Tech SIG at the Cubbery Community Center in Palo Alto. Details here:

Zvents is developed primarily in Ruby on Rails, and I can't say enough good things about it. Come find out more about the next wave in Web programming!

SDForum runs all sorts of interesting technical and business events related to software development- you can see their full schedule here.

Squidoo and TechCrunch: Experts Rise Again

When I heard about Seth Godin's new startup Squidoo, I thought, "Aha! That fits perfectly into Nivi's Trillion Dollar Matrix in the 'experts' category." I did a quick Technorati search on the status of the conversation, and discovered that Adam Marsh at EconoMeta has taken the matrix farther in interesting ways -- but pulled the experts column! Not so fast, Adam! In addition to clever programmatic methods like Squidoo for managing and measuring expertise, there are plenty of other expert systems out there. Like Google Answers, or Amazon's Mechanical Turk, which I predict will bifurcate into two main areas of practice; one, the mass processing of tasks where non-expert humans beat computers, and two, the focused processing of tasks where human experts beat computers.

But by focusing on point examples like the Turk or Squidoo, we're missing the hundred-billion dollar present reality that should be smacking us in the face:

Media is about expertise.

I regularly hear about new startups like Squidoo from TechCrunch, which is one of my key personal filters for what matters on the cutting edge of the web. Is Mike an expert? Of course he is, and the fact that TechCrunch has more traffic and better adoption curves than a lot of supposedly hot search startups speaks volumes about the world recognizing the value of expertise.

Talent will out.

The computer era has made clear that talent is much more unevenly distributed than anyone would have thought. In 1975 or 1985, were all the best managers in the world Harvard and Stanford MBAs? No. No. In 1995 or 2005, were all the top engineers in the world at blue-chip corporations and top-tier universities? No. No.

Expertise requires talent.

Let me re-assert Barry Diller's claim that there's only so much talent in the world. Diller's rule of talent (you can probably find the Web 2.0 podcast somewhere, but it parses to "there are few undiscovered geniuses in closets anywhere") means that the best user-created content is a transitory phenomenon. A lot of what Jeff Jarvis refers to as "user created content" is, I would assert, a snapshot of the emergence of talent in new and unexpected places, like blogs. But these people usually don't stay "users." Talented folks who were unevenly distributed vs. the expected structures of recognition, wealth, and power, have a tendency to rapidly move into expected positions, or start reorganizing the structures to reflect new realities.

So we end up with new media, containing new experts, who continue to act as an incredibly valuable relevance mechanism for data.

Hats off to Seth. And hats off to Mike, whose user blog --> expert media site should in no way be thought of as any less complex, cool, or important than startup entities in a different, algorithimic column of the trillion-dollar matrix.

My next post will hopefully not be so long in coming: Thoughts on how it's easy to get stuck between media and practice, drawn from a couple conversations I've had with two guys who are doing both.

Mike: media & practice
Andy: media & practice

My media is here at Onotech; Zvents is my practice.

Tribe Founders Launch Startup Aggregate Knowledge

Paul Martino and Chris Law, both founders of Tribe Networks, have been hanging out at the NSV offices and working really hard on their new startup, Aggregate Knowledge. They have a recommendations web service, WSRelater, live now on the web. I've played with it, and it's awesome. There is some impressive math behind the engine, and the number of cool things I can think of to do with it is rather long. Keep them on your radar screen.

Update: Google Responds, Urchin Works

I sent out an email to the team last night:

Hallelujah! We've got data!
After 49 empty hours, we've got data through 1pm today. Urchin works again!

Google customer service both called and emailed us yesterday as well, apologized for our troubles, and offered to refund money to us. I told them that a refund was nice, but what's really important is that Urchin works. To me, that means our website data consistently available with no more than 2-3 hours' delay.

The big issue is not one startup's angst about its temporary lack of web stats. The big issue is whether it's possible to reliably deliver complex software over the Internet at all. Google is the biggest player in this space, with vast resources of engineering, bandwidth, and servers, but everyone from Salesforce.com to Zvents is attempting to deliver significant software functionality remotely over the Internet.

We are asking our customers to bet their businesses on our ability to deliver. "Flaky but free" simply does not cut it any more. How many of you out there rely on Yahoo or Google email for business-critical communications? What if it went down tomorrow, for 48 hours? What if Salesforce.com stopped working for 48 hours? What if Ebay or Adwords stopped working? Marketing campaigns, communications, and commerce grind to a halt, and real damage is done to real people and real businesses.

Can this really work? The whole premise of the next generation of Web startups is that it can. Zvents is planning on embedding its calendars and serving events into a lot of other people's websites, such as early adopter LinkSV. If we go down, they have no events calendar on their site. As we grow, that customer reliance on us will grow to tens of thousands of similar sites, and if they have no data for 48 hours, there will be anguished screams that will sound very similar to my own post.

I think it can work, but it's going to require not only a business model, but a lot of grownup enterprise kind of thinking, with SLAs and obligations and recompense for failure clearly laid out in contracts. Otherwise, the level of trust simply won't exist to create the very cool future that we're imagining and building right now.

Google: Start acting like a real business or you're doomed

Zvents is an Urchin hosted customer. Yes, customer, as in we pay Google $200 per month to provide our web stats to us. We have been very happy with the quality of the tools that Urchin provides, but intermittently unhappy with the time lag it takes or our web data to appear, and the occasional gaps in our data (we are still missing two hours from last Monday). Google has been kind of OK on the customer service front.

Until today.

I don't know how many other people like us there are - people who paid for hosted Urchin before today's announcement. But I bet that all of them are pissed.

Here's what happened:

I tried to log in to Urchin. Surprise! Urchin now resolves to Google analytics. I typed in our login and password, and I was informed that this account "has to be validated". As it happens, this is a special account just for Urchin and I don't havedirect access to the email address - a not uncommon problem in any organization larger than, say, one. After a few hours of fumbling around and swapping phone calls and emails, I get one of the other guys to send me the validation information.

Next surprise: Because Google is switching to some damn unified login system, whenever I use the login for this account, it messes up any other Google logins I have. Apparently there's a way to pipe this login over to my Gmail account, but it's another hassle-filled step that I didn't have to take yesterday when Urchin JUST WORKED.

Eventually I get in to Urchin. Urchin is getting slammed with traffic. We hear all this hoorah about Google's hundreds of thousands of servers and their ability to host applications, and all I can say is, if casual interest in a service you aren't even accepting new registrations for grinds you to a halt like this, you are nothing like ready for prime time.

Here is a screen shot of what Urchin looks like to me after 10 minutes of waiting and a couple attempts at reloading:



Not very useful.

When I finally do get in to Urchin, I have no data since 3pm yesterday. That's 24 hours without data. Zero. Zip. None. I'm trying to run my business here, folks! Hello?

It's nice that it's now free. But I was willing to pay $200 per month for a service that Just Worked. Google may be all excited about advertising business models, but there are billion dollars businesses built on charging other businesses $200 per month.

Right now, I feel like Google doesn't care about me enough as a customer to tell me that they're changing a product I pay for. They don't care enough about me as a customer to make sure that my login doesn't change, or that they at least ask or warn me before changing my login. They don't care enough about me as a customer to make sure that the re-launch of their product doesn't dramatically impact the people who are already paying them lots of money.

Google isn't acting like a real business, they are acting like an over-enthusiastic Golden Retriever puppy. Oh, they just knocked the vase off the table with their tail, but aren't they cute? Um, no. Google, grow up.

Hand-waving hippie dreams of new media

Uncharacteristically, Jeff Jarvis posted some complete nonsense on Buzz Machine yesterday. He wrote:

Robin Good puts together a few PowerPointable lines on the future of media:

Consumers become producers of content, and niche content surpasses by orders of magnitude the value of traditionally labelled commercial television and film.

The value is not anymore in the best seller or in the blockbuster.

The value is in infinite choice of content and in the opportunity for the consumer to see content when she wants it: prime time is anytime, and anytime is prime time.

Meanwhile, allow me to quote from a recent Slate review of the new Comcast NHL coverage, a piece entitled "The latest desperate attempt to top ESPN":

...if OLN's occasionally amateurish-looking production affirms for viewers that only ESPN can do sports right, then Comcast's national ambitions could find an early, icy grave...

...The network's buffoonish studio show doesn't do much to enhance OLN's credibility....

...The telecasts also gain instant credibility because of play-by-play announcer Mike "Doc" Emrick, who leaves the cliché-ridden competition in the dust—skaters don't merely pass the puck, they "feather it along,"...

...It's easy to sympathize with OLN's early struggles. The channel won the rights to broadcast NHL games...not two months before the opening face-off. At this point in its nascent coverage, OLN's basic charge is to transmit a crisp picture; at least the network's production values compare in quality to local cable broadcasts...

Unless you're so bought in to the new memes of free content from all (dude) you'll note a trend in the Slate review (and yes, they are part of the horrible MSM, in a new-media kind of way)

1) Talent
2) Production values

- "Doc" Emrick is talented and a plus
- The studio team are buffoons and a minus
- Production values are not yet up to ESPN's standard since the OLN only had two months to put the broadcast together, and viewers will notice.

Media is valuable, because our time and attention is valuable. We all choose to consume media from preferred providers, and the two mechanisms by which we choose those providers are the talent on display, and the production values around that talent. I loosely define production values as "capital investments which make the consumption of the core content more pleasant, informative, or easier". For writing, production values include good web design, non-intrusive ads, reasonable colors and fonts (get busy, Andrew!) and so forth. For audible media, production values are things like the recording quality, background noise, the level of editing to remove slow or redundant sections, and so forth. For anything visual, the list of production values begins to be very long, and even more important.

It's a cliche that no one watches home video, despite its existence for decades. The two areas where I can think of home video having real general media value are a) occasional "man on the spot" news clips (Rodney King) and "America's Funniest Home Videos". What links these two together thematically is that they're essentially random -- something either funny or important happened in front of a turned-on video camera, and the value of the content outweighed its lack of production quality.

Here's a graphic that I put together in 2001 for a strategy I was doing for a mobile phone company. Any similarity to Web 2.0 standards is strictly coincidental - I've been thinking in power curves for a very long time. As you can see, based on the empirical data of a search I did on Amazon, there's a nice inverse relationship between cost of production and number of content items.



Why is this? There are two dynamics at play. First, in any gathering of people to create something, some people are more talented than others. I participate in some small-scale local theater and performance, and when there are six people in a room, usually everyone knows that one or two of them are better than the rest. These talented people tend to like working with other talented people, and over a small amount of time, quite a sorting effect occurs. Pretty soon the talented gang invents a name, throws out a shingle, starts charging, and voila! the naescent MSM is born.

When you add money into the equation, the effect is to enhance the differentiation. Perhaps a financier made money once, sometime, in a far distant past, by *not* seeking out the best directorial, writing, performing, and technical talent for his production - but I doubt it. When real dollars are at stake, people seek out the best, and then fund them. Those who aren't the best don't get funding. From movies to music to television to video games, this is an obvious trend even in relatively "new" media like video games.

Example of the "talent self-clusters" point here, here, and here.
Example of the "financier seeking the best talent" point here and here.

Oh, ze world, she turns but she stays ze same...

I do believe that temporarily, at least, some barriers have fallen in our new world. Here I am, writing this blog. But my choice with this blog is very clear: I can either a) invest a lot of time in it to create quality content on a regular enough basis to develop a readership or b) I can get on with my real life, which is running a cool new startup called Zvents, and only post here occasionally because I feel like it, and not worry about becoming a media property. I am never going to out-do BuzzMachine in popularity or readership unless I put some serious work into the effort, which I'm simply not going to do. Therefore I will 98% remain on the "Consumer" side of the blog equation, despite my reasonable levels of inclination, ability, and empowerment to become a producer. Meanwhile, Jeff will sit on the opposite side of the equation, and take each successive step up the production values and monetization ladder which develops over the next few years, and in 36 months, it won't be possible to compare my blog to his with a straight face - if it even is now.

Jeff will have, by dint of wildly disproportonate readership, a "blockbuster" on his hands, giving lie to the baloney in his recent post.

Whenever anyone starts talking about new democracy in media, just remember that for the past 100 years, any normal person with about $50 to their name has been able to buy a typewriter and a sheaf of paper, sit down, and create a novel. For the past 100 years, being a successful novelist has been a well-paying, respected position. The fact that so few have managed to do this, despite the legions who have tried out of the near-totality of all who could have tried, indicates that there's something slightly more fundamental going on than big mean corporations holding back the little guy.

That something is the rarity of talent, the social network clustering effects of production values, and the reality that all of our time is limited, and brands are a great way to short-circuit searching costs.

The wonderment of the current age of the internet is not that the new winners will look any different than the old winners, but simply that a window of opportunity exists to make new winners at all.

The Real Microsoft Memos

Everyone is talking about the transparently leaked Microsoft memos. The version presented here (courtesy of the unchained snark that is The Register) is much more entertaining than the ones you've read elsewhere:

Gates stirs Microsoft with dramatic 'more meetings' plea

By Andrew Orlowski in San Francisco

Published Thursday 10th November 2005 00:04 GMT

Analysis Ever the master of public relations, Microsoft has always been able to figure its way out of a tight spot with the use of a judiciously leaked memo.

...in the spirit of the excellent 500-word "digested reads" offered by some of our better newspapers, we give you the précis of the latest Gates and Ozzie memos. Then we'll put the whole affair in some historical perspective.

From: Bill Gates

Sent: Sunday, October 30, 2005 9:56 PM

To: Executive Staff and Direct Reports; Distinguished Engineers; All TV, print, radio and internet news outlets; All bloggers [delete last five before release]

Subject: Internet Software Services

Microsoft has always had to respond to innovators in the software business and seize the PR initiative.

Ten years ago this December, I wrote a memo entitled The Internet Tidal Puddle in which I attempted to undo the damage caused by my book, The Road Ahead, published just three months previously.

In The Road Ahead I failed to mention the internet at all. The same month, Netscape had floated on NASDAQ, creating the largest ever first day gain. Clearly, there was a puddle in the road ahead. So in my memo I warned that we could either step round the puddle, or step right into it, and risk being drowned.

In 1995 none of our products made use of TCP/IP. Now, a decade later, I can safely say that many of them do.

In 1995 I warned that we risked losing the mindshare of a new generation of internet software developers who chose to shun Microsoft APIs completely. Now, a decade later, I can say that close to 100 per cent of Internet-born viruses, Trojans, rootkits and worms use Microsoft APIs exclusively.

However, to lead we need to do far more.

In order to execute on this opportunity we must look as if we're acting quickly and decisively. Recently competitors [Steve, chair down. DOWN.] have gained massive public attention by rolling out products that remain in beta for many years. This is unacceptable. Products that remain in beta for many years should be first and foremost, in the public's mind, Microsoft products.

For three months Ray Ozzie has been sitting in his orgone accumulator devising a strategy to ensure we can execute. I've attached a memo written by him that I think strikes the right balance between pre-announcement and non-delivery. I'm happy to share it with you.

Bill

Go read the rest here.

NumSum, Nivi, and a Trillion Dollars

Nivi just posted a response to some of my stuff on social web. We also had a really interesting conversation over at Bessemer with James Cham on this subject the other day. Nivi's matrix is empty and hard to play with, so I put in in NumSum, which is Steve Yen's latest project:



Play away.

SDForum Search SIG @ Microsoft Thursday

John Battelle and an interesting panel. I will be here:

Scoble asks what you call 'em? Raw Potatoes!

Scoble asks what you call things like Flickr, Microsoft Gadgets, Google Maps, Amazon Affiliate parts.

Quote:

I’ve been struggling to communicate with others what the new Internet ecosystem is made up of and it hit me a few minutes ago.

They are Internet Connected Components.

Hear me out.

When you go to Kayak Buzz or Zvents, what do you see? Two ICCs. One is a Google Map. Another is a AdSense bar...

...I think we need a non-branded name to generically refer to these things. What do you think?

Here's my proposal: raw potatoes. Why? Because they're what you mash up.

If you call 'em rawpotatoes (no space) you can create a nice searchable tag/phrase that doesn't currently mean anything else. When's the last time you saw a search result this empty?

Hats off to Chris Law for reserving the URL already. I'm sure he'll be happy to build it as an open community resource for Web 2.0 just like he's done with WSFinder.

Yahoo clarifies Maps API

Vince has posted over at YDN about the Yahoo Maps API:

I wanted to take a brief moment to clarify our terms of use for Yahoo! Maps, as many folks have contacted us asking whether they could use the APIs on their site.

Our position has always been to allow usage of the Yahoo! Maps APIs free of charge for non-commercial use, as well as commercial use granted on a case-by-case basis. This is defined in our FAQ which also has instructions for how to contact us should you want to seek permission for commercial use.

In general, if you are displaying mashups featuring Yahoo! Maps on your site or application and you make your stuff available for free to users, you’re welcome to use the Yahoo! Maps APIs. This is true even if your site is supported by ads -- even ads from other vendors.

But while I've got you thinking about monetizing your site, allow me to make a shameless plug for something else we’re doing at Yahoo! -- take a look at the Yahoo Publisher Network BETA. Revenue, relevant ads, no hassle. Easy. Sign up today for your YPN beta invite.

Check out my follow-up post at the Zvents blog for the rest of the story.

The Zvents team is now going to get back to work building product. Google, Yahoo, Scoble, we love you all. Peace and APIs.

Blogosphere Speaks, Yahoo Responds (Cool!)

Yahoo launched a cool new maps service today, which got great reviews, tons of buzz, and complaints about commercial restrictions.

They also got a "Yahoo Maps are Doomed" post from Robert Scoble, in which he used my company Zvents as the poster child for Google's ongoing dominance.

Not so fast, Robert. More on this tomorrow, but I just got off the phone with the Yahoo Maps team, and they said that tomorrow they will be removing the "non-commerical only" clause from their TOS, and that Zvents, as a commercial site, is "golden" to start using their APIs.

Yahoo, you rock!

Robert has raised some fascinating points in his post. I'll address those on the Zvents blog tomorrow morning. Watch that space.

The Sincerest Form of Flattery...

This post appeared on Craigslist the other day. It's funny enough that you should click through to Flickr to read the details:



We thought it was awesome, so we created a proper home for it on the wall at the Zvents office:



Humor aside, this raises a key point about Web 2.0. The same factors that mean that "two guys and a router" can build a business like Facebook in ten days and have it take off, mean that two more guys and another router can build a business and compete with your cool new business.

There are only three ways to get to market differenation: Technological advantage, network effect, or marketing-driven aggregation of eyeballs. There are great examples of success in all three categories; but there are no examples of success that didn't pull off one. And we all know that the first two are waaaaay cheaper than the third.

It's a marathon, guys. Keep running.

Comment Hurdles, Zipless Fucks, and the Extended Conversation

At the end of a great review of Zvents (Thanks!!), Zoli Erdos points to my blog and says,

"the last link happens to be Ethan Stock, Zvent CEO’s blog. Hmm… I don’t see any way to leave a comment or trackback – what happened to the “conversation”, Ethan?"

I thought about posting a snarky comment on his blog, but this is what faced me:



Yuk.

Zoli, it was easier to come here and write this post than it was to face that. You're linked to me, I'm linked to you. That's a conversation.

What's going on here, is the empirical demonstration that anonymity and conversation are inimical to each other; and that in order to participate in a community, you must commit and invest resources. In the classic phrase of Erica Jong, there is no 'zipless fuck' in the Extended Conversation.

Zoli has a blog, and I read it. I have a blog, and he reads it. We can converse through links and references. Each of us has put some time and effort (commitment and resources) into building up this presence in the community, and cautiously, bit by bit, each of us is willing to invest time and energy and possibly a RSS subscription or a delicious bookmark to the other.

Comments on blogs were conceived in a simpler time, or with a simpler mindset, that said that anyone should be able to drop in and speak their mind - the zipless chat. But spammers infected that ideal, as sure as hepatitis and AIDS ended Erica Jong's free-love early 70s.

This isn't a sad story -- community and relationships and investment still yield great results, and if we want unexected, serendipitious conversations with each other, the best way to do that works even better in this era of the Internet -- face-to-face, the old-fashioned kind of communal commitment.

What Kind of Society Are We Building?

I've spent much of the last six months heads-down building Zvents. I am incredibly excited about the opportunity to build a great new product that will, in some small way, change the world and make users happy. Users are these slightly foreign beings to us product-builders, and sometimes it's hard to get inside their heads and understand what their real concerns are. When I read something like this New York Times article on Wal-Mart's HR practices:

"...The [board] memo acknowledged that Wal-Mart, the world's largest retailer, had to walk a fine line in restraining benefit costs because critics had attacked it for being stingy on wages and health coverage. Ms. Chambers acknowledged that 46 percent of the children of Wal-Mart's 1.33 million United States employees were uninsured or on Medicaid.

Wal-Mart executives said the memo was part of an effort to rein in benefit costs, which to Wall Street's dismay have soared by 15 percent a year on average since 2002..."

I am stunned. How can any user be thinking about movies, or bands, or Little League games, when their kids don't have healthcare? Will they ever use Zvents? Not likely. Wal-Mart is the largest U.S. employer and regularly held up as a paragon of new business practices. Why isn't Wall Street worried about the 46% of Wal-Mart kids who don't have health care? Even in the most narrow commercial view, aren't those kids the future customers and employees of the Wal-Marts of the next generation?

Henry Ford single-handedly created the 20th century social compact in 1914 when he unilaterally raised the minumum wage of his workers to $5 per day, reasoning that if his workers couldn't afford to buy the cars they were building, the great hamster wheel of capitalism might stop spinning:

On January 5, 1914, Henry Ford announced a new minimum wage of five dollars per eight-hour day, in addition to a profit-sharing plan. It was the talk of towns across the country; Ford was hailed as the friend of the worker, as an outright socialist, or as a madman bent on bankrupting his company. Many businessmen -- including most of the remaining stockholders in the Ford Motor Company -- regarded his solution as reckless. But he shrugged off all the criticism: "Well, you know when you pay men well you can talk to them," he said. Recognizing the human element in mass production, Ford knew that retaining more employees would lower costs, and that a happier work force would inevitably lead to greater productivity. The numbers bore him out. Between 1914 and 1916, the company's profits doubled from $30 million to $60 million. "The payment of five dollars a day for an eight-hour day was one of the finest cost-cutting moves we ever made," he later said.

There were other ramifications, as well. A budding effort to unionize the Ford factory dissolved in the face of the Five-Dollar Day. Most cunning of all, Ford's new wage scale turned autoworkers into auto customers. The purchases they made returned at least some of those five dollars to Henry Ford, and helped raise production, which invariably helped to lower per-car costs.

We're a tiny little company at Zvents, and soon we'll have a health care plan in place for all our employees. Not just for the highly-paid engineering types, but everyone. Wal-Mart can do better, and if they don't, and if Wall Street doesn't lengthen its short-sighted view, the hamster wheel might start slowing down.

Update on Cringely's Bad Math

Following up on my highly skeptical response to Cringely's "the sky is falling" web-hosting post, Brad Gibson, who designs data centers for a living, dropped me an email suggesting that I check out his blog. Here's a sample of his analysis:

Cringely overstates the actual (as opposed to the name-plate rated value) power consumption of the drive arrays by at least double. He also gets muddled up when he discusses data center floor space; mixing up total square footage with the square footage of the connected load space. He uses an incorrect approach to calculating "necessary" space to support "equipment" space. He makes some incorrect assumptions about the types of drives used in many of the arrays. The Cringely argument also completely overlooks load diversity in a data center.

Check out Brad's excellent piece here.

Brad's post also pointed me to the coverage of the Cringely piece on Memeorandum.

Three Postulates: The 'Long Tail' <> User-Created Content

At this point, anyone who is reading this blog has seen Chris Anderson's 'long tail' graphic so many times that I can *not* post a link to it, and say, "the yellow kind of clashes with the red/pink background, don't you think?" and you'll all be able to bring up a mental image in your head.

Let me tweak your memory further. Remember Chris' examples? The two key ones that stick in my mind were Amazon's catalog, with that fantastic quote to the effect of, "half the things we sell in a given day, we only sell one of"; and Netflix's movie rental business, where 'infinite shelf space' enables even the most obscure movies to add to the bottom line of the business.

Sure, Ethan. So what's your point?

Well, may I humbly point out that pretty much none of the items for sale on Amazon, or for rent on Netflix, are user-created in the sense that all we Web 2.0 junkies think of user-created content. None of them are user-created in the way that blogs are user-created. They're all commercial movies, or commercial books or toys or power tools, that may be rare and obscure, but were created by some professional organization with commercial intent. You do not see things like Harry Potter Fan Fiction making up even a small part of Amazon's long tail. Why is this? Not to be mean or anything, but intellectual property issues aside, there's an obvious grammatical error in the first nine words of the top-rated Harry Potter story on that site, and it doesn't get much better after that.

Let's look at another bastion of user-created content, Slashdot. Slashdot works like this: It has a home page, and a set of topical sub-pages, which are run by editors. They choose what stories make it on to those pages, and then all and sundry comment on the stories, and moderate each other toward nirvana or oblivion. Note that the stories are a) written by someone who is not a Slashdot reader and b) edited / selected by someone who is not a Slashdot reader. Slashdot readers then, with either great enthusiasm or great skill, but rarely with both, comment on the articles, adding a valuable corona of information around the core star of the article.

Are crowds wise? Sure they are. But we all agree that committees are stupid, and isn't a committee just a small crowd?

Ethan's First Web 2.0 Postulate:

"The most successful new businesses will not be built on user-created content, but will utilize it extensively."

Google is a perfect example of this postulate in action. They crawl the web extensively, and aggregate everything they can find. They then apply relevance algorithms such as hubs + authorities to decide which pages to serve up. These relevance algorithims are, as several people have recently twigged to, social software -- aka user-created content. (Note: I blogged about this in late June -- I guess I need more readers!) However, the content which we are seeking via Google - the stuff which the users are ranking through their links -- is almost always professionally created content with commercial intent, although much of it comes from the very long tail of obscure professional content.

Interesting, huh? Google takes on the one hand, an index of all the content out there (most of the useful stuff professionally created) and then adds a relevance algorithm based on what a bunch of users think of that content, and voila! market-leading search. Then they add a bunch of professionally created ad content, and add a relevance algorithm based on what a bunch of users actually think of those ads, and voila! A $106 billion in market cap.

Does this story sound vaguely familiar? How about Amazon's professional-content store, with a heavy wrapper of user opinion and review? How about eBay's professional content marketplace, with a heavy wrapper of user opinion (seller and buyer ratings) plus user-specified value algorithms in the form of final auction prices? I think I see a trend.

Ethan's second Web 2.0 postulate:

"The number of users who will usefully comment or vote upon a topic is one to two orders of magnitude greater than the number of users who will usefully create content for that topic."

Since we all know about Metcalfe's law and its new-meme supercharged cousin Reed's law, which Fred Wilson has been talking about, it's clear that if we want to build highly scalable value in the network, we might be better off focusing on capturing user opinions at scale, rather than capturing user content at scale.

Huh. I kind of think as I write these (snarky readers will respond, "We can tell, Ethan," while super-snarky readers will respond, "Not that we can tell, Ethan"), and I've just realized Ethan's Third Web 2.0 Postulate:

"Every time I think hard about where the web is going, del.icio.us seems right in the middle of it."

Cringely's Math Makes No Sense

I read Robert Cringely compulsively - he's a lively writer, a deep thinker, and totally plugged in to several frontiers of technology. But every once in a while (as with his July column on Skype) he makes claims that are not just wrong, but ludicrously wrong.

Here's the essence of his last column, which I have condensed to save you the three minutes I wasted reading it:

"A lot of money is being bet on a future user computing experience based on web services" but "the cost of keeping [user] data online all the time will be huge. It's an energy crisis in the making." "202 million [U.S.] Internet users..." * "a free Gmail account with two gigabytes of storage..." = "400 petabytes. That's 400 times the current capacity of the Internet Archive."

"that's really only about $25 million in disk drives..."
+ "total power consumption up to just under 10 megawatts, which at typical U.S. industrial power rates will cost about $5 million per year."

"This is the kind of planning and provisioning required to support FREE services...That's a heck of a lot of ads."

"My point here is that we're entering another period of Internet exuberance...the Internet will change even more than it has the ways we live and work. But it isn't going to come easy and it isn't going to come cheap."

Huh? Cringely just wrote that the big portals can create a kick-ass email service for *every single internet user in America* for a disk drive cost of $25 million, and an annual power cost of $5 million. Generously speaking, let's say that the total capex would be $125 million and the total annual opex would be $25 million (5X on both his numbers). That's a capex per user of 62 cents, and an annual opex per user of eight cents. And he says this is a problem? Dude, that's two postage stamps!

There are three great systems of analog user-created content creation and distribution in place today. They are the U.S. Mail, the phone system, and the camera/photography/film infrastructure. Voice, text, and images, analog-style. To give Cringely as much slack as possible, let's look at the smallest and cheapest, film photograpy. I called a knowledgeable friend, who told me that in the United States, there are about 25,000 minilabs for processing photographic film, and that the average cost of these labs is about $200,000. That's a capex of $5 billion. And that's just for photo prints at your local Walgreens, CVS, or Ritz Camera. If we assume one $10 per hour clerk full time to run each of those minilabs, our annual opex is $500 million. Don't forget that behind those minilabs are film plants, paper plants, distribution networks, and SuperFund sites.

By any wild stretch of imagination, the new digital infrastructure being put in place will be vastly cheaper, both to build and to run, than the old analog infrastructure that it's replacing.

The value of that infrastructure, compared to its cost, is incredible: Cringely carps that $30m is "a heck of a lot of ads" the day before Google announced that they'd just sold $1.05 billion worth of ads in their last quarter alone.

Wake up, Robert, and smell the bits.

TechCrunch BBQ 3: This Dog Got his Dinner

Michael and Keith threw a great party for the 3rd TechCrunch BBQ. It's kind of amazing how fast things are moving out here - Just as an example, I met both Kevin Burton and Jeff Clavier for the first time at the second TechCrunch BBQ, and I've seen each of them so many times in the past few weeks that they're starting to feel like old friends.

I have been deeply conflicted about the whole Web 2.0 buzz. I think that a lot of us who have just been trying to build great products for the past 10 years feel like a dog that's been patted and kicked intermittently, and we're not quite sure if it's time for dinner or a beatdown. So everyone is working their tails off, but mixing it in with ongoing self-mockery and self-reflection. At my first startup ('96), we took to calling ourselves "the last software company" as we watched a bunch of "web" companies take over, take off, and fall off a cliff. Now Zvents is running as hard as it can, while looking around at some of the fluff and saying, "maybe we're... kind of... a web 1.3 company?" The Web 2.0 conference made things worse, not better. It was crazy -- at $2900 a head or whatever they were charging, it drew some people who think too much about money, and not enough about building stuff. People who needed to be talked to in big ballrooms by reassuringly profitable, public, branded figureheads, before going out to open-bar parties hosted by a bunch of cool 19-year-olds and saying, "so this is what '99 was like."

No, actually. For it to have been like '99 would have taken a lot more MBAs, a lot more New Yorkers, a lot more 24-year old sociology chicks making $150K in high-tech PR, with a vast fog of general cluelessness about technology AND business overlaying it all. Thank God it wasn't anything like that bad.

With all that angst rattling around in my head, Michael's party tonight was amazingly refreshing. It was pretty much entirely populated by real, genuine, Silicon Valley people. Startup people. Big company (Google, Yahoo, IBM, Ebay) people. People who've built stuff, and done stuff, and made things happen. People who understand that you have to cooperate to make standards, and compete to make great products. People who know that elegant solutions to computationally hard tasks matter, and that compelling end-user experience matters too. It felt very real. There are some great products being built, and we got some fascinating peeks at What is Yet To Come.

I feel much better about this whole Web 2.0 thing after tonight. This dog got his dinner.

P.S. A few pictures:




See Kevin's Feedblog for details on Dave Winer's $20

This post should probably be tagged techcrunchbbq or techcrunchbbq3 or somesuch.
It's the Tag Guess Dance, aka Bubble Up from the Bottom, Baby!

Kosmix.com, Vertical Search for Healthcare, Launches

All of Web 1.0 continues to be reinvented around the search paradigm. Healtheon/WebMD, watch out: Today, Kosmix.com launches to provide healthcare search. I know two very smart search engineers there, Jason Zien (from IBM's WebFountain) and Ram Subbaroyan (from Inktomi). Kosmix is founded and funded by the Junglee/Cambrian nexus, and I expect them to kick some serious booty.