Graeme Thickins' Takeaways

Web 2.0 Is Going Beyond the Consumer Market: There's no doubt these technologies will be increasingly applied within enterprises, and that's a good thing. The big vendors see this and were all over this venue - exhibiting, presenting, and otherwise.

The VC Investment Model for IT/Software Is Changing: Because new Web 2.0 ventures don't need much money to get going, the VC business is being somewhat disrupted. A new style of early-stage VC firm seems to be gaining. It's becoming obvious that the big funds don't fit in here. These days, a mere $250-750K is all that's needed to prove assumptions, one VC said. And angel investors are newly energized.

Marketing Is Rising: And one of the next best quotes I heard was in that same panel, from VC Jeff Clavier of SoftTech: "We need to start seeing real money spent on marketing." The signs are good. Everywhere at Web 2.0 Expo, I saw evidence that a large part of the conversation, and in fact the event's program, was about marketing. Kudos to the producers for that!

SEO Is Rising: A corollary to the above is that SEO, far from declining as one noted blogger/entrepreneur posited recently, is becoming a major focus in the marketing world - hand-in-hand with skills in social media. This is the new sandbox for marketers; and that is exciting. Widgets are emerging as well - another important and related marketing medium.

Another Reason Not to Get Cocky: Lest we think that we as Internet users are so much in control today, consider this: there are only one billion PC users, while two billion people worldwide use mobile phones. The Web is far from universal yet and certainly it will be some time before the advances that we call Web 2.0 become widely used worldwide - on desktops, $100 laptops, cell phones, or whatever. Maybe next year's Expo should be in China or India?

Offline Web Apps Are a Next Big Thing: The demo of Apollo, by Kevin Lynch of Adobe, was awesome - even though it is still in alpha. And Etelos, another event sponsor, had a great way to position their new offline offering: "Apps on a Plane." Who wouldn't like that?

Geeks Are Not Good Presenters: Please, conference producers everywhere - get your presenters and panelists ready to go up on stage in a big room. And, for that matter, the marketers and CEOs too. The vast majority of the presenters I heard were sadly lacking in presentation and communication skills. I guess there's a reason why conferences like DEMO work hard on this. Take a lesson. [Ed: I presume Graeme didn't mean me in the Hybrid Design panel.... hmmm, ok maybe he did!]

Sean Ammirati: What's going to beat Google?

Dave Hornik had an interesting post this week, discussing how the venue of an event shapes the things attendees take away. At the scale of this conference (I heard 16,000 attendees?), a lot of the value comes from networking interactions. The show didn’t disappoint, I arrived with a stack of cards an inch thick and they were gone in a day.

In all of these interactions, one theme almost never failed to come up --- how Google related to their project or company. The humor wasn’t lost on the audience when John asked Eric Schmidt about the antitrust concerns that Microsoft and AT&T have raised and Eric joked “Wait, Microsoft has concerns…” [about anti-trust]. Yet, joking aside, it is obvious that Google is the 800-pound guerilla now. That said, I’m confident that just like with Microsoft and IBM before, there are vulnerabilities to Google’s business model. It will just take another entrepreneurial company or project to leverage an unanticipated paradigm shift and reveal those weaknesses. Therefore, the thought I left the Expo with is … what’s that concept? Only time will tell, but there is a good likelihood the founders were walking the halls at Moscone West this week!

Emre Sokullu's Thoughts

Web 2.0 Expo was excellent for networking and meeting great people. The two technologies that impressed me most were: 

• Live texting feature during the Ignite event, powered by Mozes;
• Potenco, a solution to the big energy problem of under-developed countries. Imagine this being used with OLPC - these innovations could revolutionize our age.

So what was different about this web 2.0 conference? It was the 3rd Web 2.0 Conference run by CMP/O'Reilly Media that I've been to (I followed the 2004 Web 2.0 conference virtually). Check out my previous wrapups of the Web 2.0 Summit and the Web 2.0 Conference in 2005. The Web 2.0 Expo was the biggest of all 3, both in terms of the venue and the number of people attending - estimates ranged from 10-16k. It also had more of a developer focus, although there were plenty of business people too. More on that in a minute.

From a product and web technology perspective, there were a lot of enterprise or SME focused startups and products on show - particularly in the Expo Hall area (The Land of the Booths). As I walked around the plentiful booths, it seemed at least every second product was for enterprises or SMEs. This was definitely a change from the previous conferences, where consumer-focused products dominated and usually stole the limelight. I've already profiled Bungee Labs, a next generation web development platform that impressed me. A few other newcomers that caught my eye were Egnyte (a collaborative document sharing app), Vidoop (a security solution that featured in the LaunchPad), and robotreplay.com (a site analytics service). That's naming just a few, and I didn't manage to see all of the companies on show (check the R/WW Author wrapup of the Expo, in the next post, for more standouts).

In terms of the sessions and panels, the Web 2.0 Expo was overall a success - with enough good sessions about new technologies and interesting Web 2.0 issues to keep people happy. There were of course some sessions that were disappointing - e.g. some were just product pitchs, or perhaps the discussion between panelists didn't quite pan out. But that happens in every conference. There were unfortunately a lot of issues with the WiFi, despite the best attempts of the organizers to fix it. Another disappointment was that the Expo had no closing keynotes on the final afternoon, and the main booth area was also closed at around midday on Wednesday. It seemed kind of a letdown to only have panel sessions running in the final afternoon.

Photo: Scott Beale / Laughing Squid

But turning to the positive, the Expo had a great mix of people attending - developers, designers, business people, VCs, bloggers, etc. So unlike the Web 2.0 Summit at the end of last year, this crowd was diverse and more focused on the technology aspects. I also got the impression that more 'mainstream' people, from corporations or small/medium businesses, were in attendance - which is a healthy sign. And that was really the aim of the conference organizers, encouraged by the different tracks of sessions (one for design, one for strategy and business, etc). The only complaint I heard about the tracks was that some of the developer community who already knew about web 2.0, did not find enough detail in their sessions to satisfy them. But on the other hand, I also spoke to a company that wasn't familiar with web 2.0 whose developers were enjoying the tracks (this was after the first day). So I guess you can't please everyone, but overall the (admittedly few) sessions I attended were interesting. I actually spent most of my time networking/booth-hopping and in the press room, which in my case was the only place I could connect to the Net.

Highlights

Other features of this conference I enjoyed:

The Expo Hall was great, with a good mix of startups and more established vendors. Strangest sight? The IBM booth, with IBM staff milling around in red shirts (not IBM blue) and doing on-the-spot classroom sessions with microphones and videos. That strategy seemed to be pulling in the punters (many willingly plopped themselves into the IBM chairs and listened to the lessons)... but as a fellow blogger quipped to me yesterday: "I don't go to web 2.0 conferences to listen to IBM". No disrespect intended, because IBM is doing some great things now with web 2.0 technologies. But the real excitement was to be found, at least for me, in the many startup booths - e.g. CambrianHouse's booth was excellent, with a kind of tropical theme going (see pic below). All in all, I enjoyed wandering around the booths and getting demos from the many passionate technologists waiting to gush about their babies.

Photo: Scott Beale / Laughing Squid

The Ignite show on Sunday evening was awesome. These were 5 minute quickfire presentations on a particular theme or product, mostly focusing on stuff happening on the edge. I really enjoyed them, precisely because they were not the usual 'product pitches' - they were more like 5 minutes on whatever technology passion the presenter had. 

The presentations ranged from the more traditional - for example Ryan Stewart (who was a last minute addition, so he had little prep time) did a fantastic 5 minute summary of RIAs - to the bizarre, e.g. the guy who built a web app at the South Pole or the guy who discussed how his hobby of bee-keeping was a good example of the web 2.0 "hive mind"! My favorite presentation at the time was the 'open source hardware' story about 14 year boys making Lego guns at instructables.com - although given that the Virginia Tech tragedy happened the day after, it is in retrospect more than a little disturbing. At the time it came across as more humorous than creepy; but thinking about it in the aftermath of the tragedy, it certainly is a somewhat disturbing use of technology to make guns out of lego. 

In any case, I loved the Ignite sessions. Especially as the audience had the opportunity to vote for the best presentations, using text messaging on mobile phones. Big credit for the concept has to go to Brady Forrest of O'Reilly Media, who came up with the idea and has been running Ignite in Seattle recently. I will delve more into Ignite in a follow-up post.

The Web2Open was a great addition to the conference. This was a free event where developers and designers discussed the tech issues of the day. I must admit I didn't have time to attend it (other than poking my nose in a couple of times), but I did hear second-hand that the Open satisfied web 2.0 technologists and the 'in crowd'. Props to Tara Hunt and Chris Messina for organizing this.

Conclusion

When it comes down to it, Web 2.0 Expo was a success in my eyes, despite the technical glitches with WiFi and last day scheduling oddities. From a networking standpoint, it was a great crowd of people to mix and mingle with. The cliche is that the best part of a conference is the discussions in the hallway - and in the case of the Expo it was discussions in huge cavernous hallways that were usually teeming with people. I got the feeling that others enjoyed the conference, although it's hard to generalize. The evening parties were good too, with the Netvibes event on Monday probably the highlight.

If you attended the conference, please add your thoughts on it below.

All photos in this post are from Scott Beale / Laughing Squid, who has a lot of great Web 2.0 Expo photos on his blog laughingsquid.com. 

The rest of the day, I found myself reflecting a lot on the presentation and then - as luck would have it - I ran into MySQL AB's Marten Mickos and Zack Urlocker (Executive VP of Products) at a happy hour that evening. I ended up talking to them about their business over a few beers and concluded that they are a great lesson for any website looking to scale audience.

Back to the panel now. First of all it is important to point out that both these organizations operate with the intention of making money. This obviously isn't true for every one of the over 140,000 open source projects listed on Source Forge. Both companies have more than a hundred people contributing to the product as paid employees and clearly are oriented towards generating value for their shareholders. In fact, John attributed being a 'commercial open source business' as one of the keys to SugarCRM's early success.

What is interesting is that they have been able to build these businesses with so much less capital than previous closed source applications in their space. John estimated that SugarCRM has achieved the same market traction at this point that Salesforce.com did in the first 3 years, but with about 20% of the costs.

Creating an Architecture of Participation

They have done this by creating what both panelists called an architecture of participation, which they built into their business strategies. John covered six components to that 'architecture' that made participation straight forward with SugarCRM:

• Extensible Project - SugarForge: a site developers can go to and create their own complementary projects (currently over 8,000 developers have participated)
• Easy for anyone to participate (easy to download, share ideas, etc...)
• Others Can Profit - Sugar Exchange: a site for people to create extensions to sell their wares
• Transparency: for example the support forums are completely public, whether you use the software or are just interested in it
• Access to the code (it is open source after all)
• Easy to purchase the PRO edition

SugarCRM has had 3 million downloads of their application and it has been translated into 50 different languages, so they are obviously doing something right.

MySQL has an even larger audience receiving approximately 50,000 downloads a day of their application. Marten framed the significance of this in an interesting way. He pointed out that their closest (closed source) competitor has fifty-six thousand paid employees. However, every day he has fifty thousand passionate new users. He quipped that while it is easy to disregard these as mere amateurs, "Noah's Ark was built by amateurs, Titanic by professionals."

Still Have Sales People, but ...

While this model certainly has led to some interesting distribution and keeps the cost of sales low, it is important to point out that both organizations have executives responsible for sales. In fact, MySQL has over 70 sales professionals. The difference is they are focused on selling consulting services and support, instead of proprietary software. While MySQL doesn't disclose their revenue, it is about equally divided between services, support and revenue from OEM partners.

Conclusion

The Web 2.0 conference has been organized under the theme 'web 2.0 is _____' - with everyone contributing something different in the blank space. In fact, the conference organizers provided conference t-shirts with 'web 2.0 is ____' and gave all of the attendees markers to fill in our answers. Many of the answers filled in centered around collaboration with users and efficient distribution.

What I found interesting after listening to the above presentation is that not only can web 2.0 entrepreneurs build sites on top of open source projects (like MySQL - L.A.M.P.), but they can actually learn a lot about promoting their services from these products. So the next time you're in a marketing meeting trying to figure out 'distribution', an interesting brainstorming question is -- what did mySQL or SugarCRM do?

Why are Widgets Popular

Dion opened by unpacking the fact that the web is becoming much more of what he characterizes as a DIY Phenomenon. This cultural change is what he credits as the driver for the popularity of widgets. Specifically the DIY ethos on the web has four components:

Key Aspects of Widgets

Based on this change on the web, there are three key aspects that every widget should include, according to Dion:

• Supreme ease of consumption and distribution;
• Connect to their underlying sites to provide value and control;
• Have business model baked deeply into it (Driving site traffic, content consumption, advertising, etc...).

Business Case

Finally Dion got to the cornerstone of his talk, touching on three components that should be included when making a business case for building widgets. He then pointed to two examples of very successful widgets. The components of your business case should include:

• Getting your content on millions of other pages instead of just your own site;
• Letting users broaden your distribution globally 24X365 for virtually no cost;
• Turning your applications into open platforms and foundation of dozens, hundreds, or even thousands of other products.

While these all seem valuable in theory, it is important to look at widget examples that have actually contributed to an organization's goals. The first example, not surprisingly, was the YouTube Video Badge. This is an interesting example, because as readers of R/WW know, the viral popularity of a YouTube badge on MySpace and other social networks was a key component of YouTube's growth. This has resulted in YouTube becoming one of the most popular sites on the web.

Based on this popularity, Dion examined the social nature of the YouTube widget. There were two components which made this widget so easy to share and ultimately led to it being so popular:

• At the end of every video, there is one-click interaction to share the widget:
  
• Additionally, on the YouTube website itself, the code is shown on the site; allowing users to copy and paste the html code required to put the YouTube widget on whatever other site they want.
  

I would add that the widget was something that users of MySpace and other social networking sites wanted badly. However, once you have built a widget that satisfies a need for your users, it certainly is interesting to reflect on YouTube for ways you can make that widget more social.

Dion's next example was interesting because I wouldn't have thought of it as a widget. He actually pointed at the Google AdWords Widget for Publishers (AdSense) as "probably the most successful widget in history." While certainly not a widget for self-expression, it certainly has created value for Google and the publishers using that widget. Specifically, this has made pages across the web a platform for Google to sell ads on.

Conclusion

As the Internet has evolved into a 'DIY Web', it is important for every site to evaluate how widgets fit into their strategy. They clearly have been critical in the successes of many web services and hopefully panels like the two covered above will help others figure out how widgets can help them.

I'm sure that many of you have additional observations and success stories from your experiences. Please feel free to share them in the comments below.

If you're in the online marketing game and are not yet hip to widgets, listen up. Two emerging Web 2.0 technology firms focused in this space have a message for you. Those companies are Widgetbox and ClearSpring, both of which presented in a session on Tuesday afternoon at Web 2.0 Expo that was billed as "Using Widget Syndication for Online Marketing and Measurement".

What is a widget? According to Ed Anuff, CEO and cofounder of Widgetbox, it's a "small piece of dynamic content" on your web site, blog, or social network page. Sometimes they're also called badges. An example of a widget is the Flash-based Flickr badge shown to the right, which is one I use in the sidebar of my blog.

Why do you need a widget strategy?

Why would you want to have a widget strategy, asked Anuff? Because they're contextual and personalized, they're social, they're visual and interactive, and they're viral. "They let you take an experience and share it with others," he said. They allow for self-expression, they provide site enhancement, and they facilitate ads and commerce - for example, a shopping cart can be implemented in a widget.

How do you launch a widget? Well, you can use Widgetbox or similar sites to feature your widget. And you should set up special SEO-optimized widget landing pages, said Anuff. Tapping blogger and social networking power users was another recommendation. The syndication network that Widgetbox has set up already has more than 30,000 domains, Anuff said (reminding us that MySpace represents just one!). The firm also has amassed opt-in lists of 50,000 "power bloggers and social networks." 

Also Widgetbox does some unique things with SEO - it claims to have more than 10,000 widget pages in Google. The widget analytics it provides integrates with your existing tracking systems. "You need to pay attention to your widget metrics," Anuff said, noting that impressions and clickthrough rates are not enough. "It's about getting that wildfire pickup we all desire," he said.

The birth of the widgetsphere

"We're witnessing the death of the portal and the birth of the widgetsphere," said the next speaker, Hooman Radfar - founder of Clearspring, another VC-backed widget syndication company. He said the old way was for portals to aggregate content for you. "But now the 'prosumer' is here, producing and consuming." Creation is easier than it was under the old portal model, bandwidth is cheaper, and we have Ajax and Flash. Radfar said that "we just need to give people the tools. Widgets are the building blocks for social aggregators." His definition of widgets: "Components that can be executed across multiple platforms without additional compilation." 

Hooman Radfar

Radfar said that widgets are a part of a fundamental change in the platform of the web - "and the widgetspace is not just MySpace. It's hi5, bebo, and many others - even Nokia is doing it."

So, what can marketers or businesses do to take advantage of this movement? Radfar said that your strategy should be to "build a virtual destination in the widgetsphere." Use promotional widgets: "You can think of them as flyers." How this differs from advertising he didn't say - except you obviously aren't paying for placement here. This is social media, friends.

How to get started

A standard size for a widget is 150 by 300 pixels. Flash is generally used for social networks, while Javascript tends to be used for widgets on blogs and start pages. There are basically three ways you can go: 

(1) build your own widgets;

(2) use destination-specific tools, such as at Google, or;

(3) go with an independent widget tools and syndication company like Clearspring.

Radfar recommends to "let the user play with your widget. Don't force them to download and post it. We've studied this and find it gets better results". 

He noted other widget galleries where you can list your widgets, such as Google, TagWorld, and Typepad. And he stressed that widgets, which are really little apps, need to be measured. Keep adapting the widget, keep iterating. Other tools are available at Netvibes and PageFlakes, he said. 

"Widgets are write once, run everywhere, and we've already served more than 2.5 billion of them," Radfar noted. In conclusion he said that a new platform is coming from Clearsping in May.

Widgetbox

Clearspring

• SSO for the web
• Simple and lightweight
• Easy to use and deploy
• Open development process
• Decentralized, Free
• People are already familiar with URLs
• User control of information
• Site specific hacks are possible - use AOL user name to sign-in.

• Platforms: Joomla, drupal, phpbb, rails, plone
• Sites: Technocati, digg, sixapart, pageflakes, netvibes, wordpress etc.
• Vendors: Microsoft, AOL, Verisign etc.

Implementing OpenID

Brian showed a demo of how OpenID works, by logging into jyte.com. He followed it up with a cool example of OpenID delegation, which showed how users can use their own site as an OpenID and delegate the sign in/authentication etc. to another OpenID provider (OP) - with just 2 lines of code. This allows users to easily customize their OpenID, along with giving them the flexibility to change their OP when they want.

David then showed an example of how to create your own OP using phpMyID. He created a new OpenID within minutes (hashing the password seemed a bit complicated though and it will take me more then 2 minutes!). He also demonstrated how users can create their own personal profile data and control it centrally, to provide the right set of information to the right relying party. By using this technique, users will not need to fill out the same sign-up form over and over again at multiple sites.

Brian then demonstrated how to install OpenID on Ruby, using the ruby-OpenID library. He suggested that all relying parties should use the standard “openid_identifier” to name their OpenID input name, to make it easy for browsers to detect and process it. The Ruby example of enabling an app to use OpenID seemed really easy.

The Phishing Problem

To their credit, David and Brian addressed the tricky phishing issue that has been plaguing OpenID. They suggested a number of potential solutions that are being worked on:

• Client side certs (browser based certificates)
• Microsoft CardSpace (IE 7/Vista)
• Vidoop (image based access code); this is really not an anti-phishing solution, but it does allows users to replace passwords with easier to use visual categories - which defeat the keyboard logging kind of attacks.
• OpenID SeatBelt: This is a new browser plug-in for FireFox and IE by Verisign. The SeatBelt works as follows:
  - The browser plugin first detects if a web page accepts OpenID authentication;
  - It then asks the user to Login to their OpenID account, so that they don’t have to login again;
  - It shows a visual indication that the login page is safe, plus the current login status of the user as a browser button in the browser toolbars;
  - In terms of usability, the Seatbelt plugin automatically fills out the OpenID field when it detects a site that accepts OpenID.

Overall it was a great, although somewhat basic, session. If you are interested in finding out further details of the session, the slides of the session with notes are available on OpenID.net.

To find out more about the system, I spoke yesterday to CEO Martin Plaehn and VP Community Alex Barnett. Bungeeconnect is the name of the actual development platform and its most striking feature, at least to me, was that it easily allows developers to utilize hundreds of APIs - from the likes of Google, Amazon, eBay and others. The web apps that result are cross-platform (running on Windows, Mac, Linux; as well as all the main browsers), so all up it is a very inclusive system in terms of web services and the underplying PC/browser platforms. The following diagram shows the high level concept:

For developers, probably the biggest appeal of Bungee is that it reduces the complexity of building web apps that hook into multiple APIs. Connected web apps that use web services from various companies (like Google, Amazon, etc) are becoming increasingly common, so Bungee is designed for this nascent environment. It will also enable groups of developers to collaborate online - so connecting not only to external web services, but to fellow developers.

As I mentioned above, Bungee hopes to make money by being a computing utility. It doesn't charge the developer to use or test the platform, only charging for the commercial services used by the end customer. It also will not get into the business of developing or deploying web apps itself - it wants to be focused on being the platform for others to do those things. Bungee hosts the 'connected app' for free, so in this respect it is attractive to its target audience of Small-to-Medium sized businesses (SMBs). Note that developers still need to comply with whatever restrictions or fees that the API providers might charge for using their data, but the dev/hosting service that Bungee provides is free - up until the time your web app starts making money.

Bungee Connect is similar to platforms like Morfik and Google Web Toolkit, in that it's a web app development platform and one of its goals is ease the process of developing Ajax applications. Alex Barnett compared it to a web version of Microsoft's Visual Studio. It also competes in a way with Adobe's Apollo and Microsoft's own next-gen web app development platform Silverlight.

From a high level, Bungee is interesting because it shows how web apps are developing into a much more connected ecosystem - based on the hundreds of APIs and web services that are available for developers to hook into. Bungee has made partnerships with the following core API providers:

But ultimately Bungee allows you to connect to literally any service that has an API or another method of hooking into their data. CEO Martin Plaehn told me that they're targeting e-commerce and CRM providers at this point (like Amazon and salesforce.com), with their next goal the search and geo-location services that are a subset of e-commerce and CRM. So for example Kayak and Zillow.com. The following diagram illustrates this:

Conclusion

This post is just an introduction to Bungee Labs - their product and its goals are very broad. Bungee Connect is still pre-beta (it goes into beta in May), so there's a lot to learn yet about the company. For further information Alex Barnett has a great post on his personal blog, that also provides links to other analysis pieces. But if you're a developer interested in building next-generation web apps, the best thing to do is check Bungee out for yourself. Let us know in the comments what you think, and whether this kind of platform will result in the next wave of compelling web apps.

As an industry analyst and someone keenly interested in where the Web is going, Bungee seems to me to auger a new breed of web app - i.e. highly connected apps that utilize a variety of external APIs and web services, from service providers like Google and Amazon. This will ultimately result in more choices for end users, and more innovation in web apps and online businesses.

The talk shifts to DoubleClick. John says that DC was in the past seen by Google as the type of advertising (banners etc) that was oppositie to Google (CPC text links). So what's changed?  Schmidt says that Google has since decided to offer a full scale set of advertising - not just text ads. So they acquired YouTube, started doing TV and radio advertising, and more. Now Google is looking to offer a single way to do all types of advertising. Since 2004, he says DC has become more targeted and offers better support tools (for publishers etc). So he says combining this with Google's technology will make "the math work" for them, in terms of the $3B price they paid for it. He finishes by saying that Google's technology does the best job of targeting, so if you marry that with DC's people and tools, that's how they came up with the $3B price tag.

John asks what areas interest Google. Schmidt says mobile space is the biggest. He says 3G and 4G will provide a lot of opportunities and that this is "a wide open space". He also notes the local space, which is a big search opportunity.

John asks: what does Schmidt think about when he first wakes up? Schmidt aboids the early morning part, but says he thinks a lot about scaling. He worries about this - because scaling requires more data centers, cash flow, people, product announcements etc. But he says the amazing thing is how early we are in scaling the Internet. He says we're just at the beginning of getting information that was previously kept in small pockets, onto the Internet platform.

Last question, John asks about data portability. Can users get their own data and e.g. give it to Yahoo. Schmidt says that Google has made a commitment never to track personal data (search history, gmail etc). He says end users wouldn't choose to adopt the services Google offers otherwise.

Next Bill looks at participators vs viewers. Some 'visits to media upload' ratios: 0.16% for YouTube, 0.2% Flickr, 4.59% Wikipedia (entry edits). Dave Sifry notes that the creation percentage - I guess I'd call it the read/write ratio! - is certainly a lot lower than the old 80/20 rule. Bill drills down into the participatory figures, which shows for Wikipedia that older users are much more likely to be participatory (35-55) whereas the younger users are the viewers. However for YouTube it is the 25-54 yr olds that upload videos - however note that Hitwise doesn't track <18. The gender breakdown shows that 76% of users are male on YouTube, but a 60/40 male/female split for wikipedia.

Predicting the next web 2.0 winner

Bill shows stats that show that YouTube went from zero to dominant market leader in just 6 weeks! It passed Yahoo and Google video search in 3 and 6 weeks respectively. Also during the 6 week early adopter phase, there were key segments over-represented: 'money and brains', 'young digerati', and 'bohemian mix'. Bill then showed some great stats about up and coming companies Hitwise has identified. Yelp is one of those poised to gain traction, along with stumbleupon and veoh. WeeWorld, Imeem and Pixo were also mentioned.

Dave Sifry is on now, with his State of the Live Web Q1 2007. He starts off by noting that his first high order bit a few years ago was on the state of the blogosphere, but there's been a huge shoft - videos, music, podcasts, etc. Technorati is tracking 70 million blogs now, at a rate of 120,000 new blogs per day - 1.5 million new posts per day. He compares to "an enormous amoeba". He says there are 15.3 million "active" blogs (21%). He also notes the mainstreaming of these technologies - and that people may not even notice they're reading a blog. 12% of the top 100 sites 6 months ago were blogs; but now he says it is 22%. Mainstream media is still at the top (NY Times etc), but there is growth in the blogs.

He looks now at the behaviors of the top bloggers. He says these are effects, not necessarily causes. He notes that influential bloggers post more frequently, on average twice a day. Whereas "magic middle" bloggers (about 3M) post on average once a day. Also influential bloggers have been at this at least 1-2 years. Finally, 88% of the top 100 is different than one year ago - i.e. it's very fluid.

In language, Japanese is now the top language - 37%. However there are undercounts: French, Korean and Chinese. English is now 33%, only one third.

Tagging has become a mainstream acitivity in blogging, according to Sifry - 230 Million tags over two years, with 37% of blog posts using author tags.

Conclusion

This was a great double-act speech and I look forward to drilling into the Hitwise data in particular some more. Technorati's data is always interesting, especially as they've been tracking the space for so long. Also Hitwise's analysis on which companies are most likely to become popular is an excellent breakthrough - and Read/WriteWeb will get in touch with Hitwise to analyze this some more.

Techcrunch20

Mike Arrington announced his new conference with Jason Calacanis, Techcrunch20. The format is twenty new startups from around the world will announce and demo their products over a two day period - and they don't pay for this privilege. The conference is set for September 17-18, 2007 in San Francisco. The companies will be selected by a panel of 20 experts.

Start Pages Busy

Two of the best personalized start page products, Netvibes and Pageflakes, have both launched new features. Netvibes Universe lets you create a Netvibes page featuring your favorite content and widgets. This seems to be the beginning of mainstreaming the service, because already more than a hundred 'branded' universes have been created, for leading brands like CBS and popular music artists. More details here. Pageflakes has also released a lot of great new features recently, under the code name Flurry. Check out their blog for more details. R/WW will be checking out both Netvibes and Pageflakes new features in an upcoming post(s), once the Expo flurry has subsided.

Expo Action

So far the Web 2.0 Expo has been very focused on explaining web 2.0 concepts to developers and designers, so it's different from previous web 2.0 conferences. I briefly met someone last night from a Canadian startup, who told me that 7 of his company had made their way to expo to find out about web 2.0. He wasn't familiar with the web 2.0 world previously, so he said the Expo has been very useful and practical. This is also the sense I got from the panel I moderated yesterday, that it is a practical conference that is more focused on showing the world how web 2.0 technologies can be used. More on this trend in later posts today and tomorrow.

I'm off to the keynotes. If the Internet access is better than yesterday, then I'll report more soon.

This theme of a global computing network continued when Amazon.com founder and CEO Jeff Bezos came on stage to talk about Amazon's latest business venture - a giant web services platform. Bezos said that S3 now has 5 billion objects stored in it. What's more, it had 920 million S3 requests on its peak day, and 16,607 requests in its peak second. Obviously we're talking about a very well scaled system here. The term Bezos used to describe it is "web-scale computing".

Photo: Scott Beale / Laughing Squid

Bezos also talked a little about EC2, which is their newest service released in 2006 (Elastic Computing Cloud). Right now EC2 is invite-only, because it is "capacity-constrained". However Bezos said that they hope it eventually becomes "demand-constrained" so they can roll it out to everyone. There is no such problem scaling S3, which is already open and available to anyone to use.

This morning I moderated a panel called The New Hybrid Designer, with Kelly Goto, Chris Messina and Jeremy Keith (unfortunately Emily Chang was sick and couldn't make it).

From left to right: Jeremy, me, Kelly, Chris (who took the photo)

• Writing semantic markup (transition to XML)
• Providing Web services (moving away from place)
• Remixing content (about when and what, not who or why)
• Emergent navigation and relevance (users are in control)
• Adding metadata over time (communities building social information)
• Shift to programming (separation of structure and style)

Since that time, the world of 'hybrid design' has evolved further. It was summarized nicely by this image Chris put together:

The talk was more aimed at designers and developers, so my role was mainly to make observations about high level trends and the panelists would drill down into more practical details. I hope it was an enjoyable session for the audience, about half of whom were designers and the other half developers (with a healthy mix of 'hybrids' of course!).

Well, I am off to roam the hallways, check out some more sessions, and also venture into the land of the booths (where startups are showing their wares).