CS73N Meeting 14 Notes: What is Web 2.0?

Started by Avron Barr, based on class notes from Thursday, 25 May 2006. Updated by Shirley Tessler, 31 May 2006.

"Web 2.0 is the network as platform, spanning all connected devices; Web 2.0 applications are those that make the most of the intrinsic advantages of that platform: delivering software as a continually-updated service that gets better the more people use it, consuming and remixing data from multiple sources, including individual users, while providing their own data and services in a form that allows remixing by others, creating network effects through an "architecture of participation," and going beyond the page metaphor of Web 1.0 to deliver rich user experiences."

Tim O'Reilly, October, 2005

Today we looked at the Web 2.0 concept that has been so widely discussed in the press and in the marketing materials of recent Internet startups. The concept originated in a discussion about survivors of the dot-com bubble at a conference put on by O'Reilly Media. Tim O'Reilly published an article in the Fall of 2005 summarizing current thinking on what "Web 2.0" might mean. (O'Reilly Media has an annual Web 2.0 Conference, and he has also trademarked the word!)

The Internet companies that have survived the crash, and many of the new Internet startups, reflect a kind of user participation, community thinking, and data integration that was not central to the original idea of the web as a "publishing" medium. (There are exceptions, like Ebay and Amazon, which are in themselves interesting.)

Software Platforms 

Software platforms, like Windows and Linux, are software systems that allow software applications to conveniently run on several different hardware platforms.  For instance, a single copy of Skype or Excel can run on any Windows PC, regardless of its manufacturer and the specific components used in its construction. (But a separate version of the software is required to run those applications on a Mac or Linux machine.) There have been many software platforms over the last 50 years. IBM's OS360 was the first major commercial software platform that opened business computing to independent software applications vendors. (Before that, all business software was sold by IBM.)

Software platforms represent several elements:

• a set of standard interfaces that software applications can use to access the underlying hardware (disk drives, keyboards, displays, networks, ...);
•  a lot of work on "drivers" that implement those standard interfaces for all the different types of hardware, and
  
• a community of applications developers (programmers) and software publishers.
  

It is their support of this community of applications developers that has distinguished Microsoft as the leading software platfrom vendor today. (Dave Winer, in a 1994 blog!, described the relations between application vendors and the platform vendor as a Chinese household.)

When the Web first started, Netscape (a Web 1.0 company) had a business plan that viewed the Web as a hardware platform. Netscape's intentions were to become the software platform for all web applications. By giving away their browser, Netscape intended to position itself as the vendor of server-side software products that would support all web applications. But soon the web server software as well as the browser became commodities, and Netscape's business plan failed.

Google is the flagship Web 2.0 company. Google started as an application running on the web--a "web service".  No packaged software product, no licenses, no periodic releases, and only one version (vs. a different Mac and PC version, for instance). At the heart of Google, and of Web 2.0 thinking, is a special-purpose database, with many applications (and application-building tools) that interface to that data.

Characteristics of Web 2.0 Applications

Here are some of perspectives that are indicative of the second generation of web applications:

The long tail. While companies like DoubleClick were way ahead of the idea of server-based applications that blended with other applications to render a webpage, their business model was (and still is) focused on the "top 2000 websites." Web 2.0 involves the whole web and stresses the collective power of small sites, and of individual users. See Gio Wiederhold's  notes on Chris Anderson's talk on the Long Tail.

Architecture of participation. Many applications get better the more people use it. Such "network effects" are illustrated by eBay, Yahoo Internet Messenger, and Skype.

Collective intelligence. While it is certainly true that eBay and Skype demonstrate network effects -- increased value to each user as more people sign up -- Google's page ranking system goes one step further. It incorporates the decisions of each user into its data, adding value based on the users' knowledge.Wikipedia is an extreme example of the power and utility of collective intelligence. Collaborative filtering ("people who bought this book also bought ..."), in sites such as Amazon and Netflix, are additional examples.

Blogging. Another example of the collective intelligence idea that has taken on a life of its own is the running commentaries that thousands of people publish on the web. These Blogs point to each other and to interesting web pages. As a result, they impact Google's page rank algorithm, for instance, which is based in part on how many times a given page is pointed to. Key technologies here are RSS (really simple syndication), which allows readers to tune into content streams, and permalinks, which allow a writers to point to a particular piece of content in a semi-permanent way.

Infoware vs. software.  The Web 2.0 leaders have leveraged their control of their data, vs. control of software interfaces to their systems (the API's that microsoft controls in Windows). To the extent that the data has value and is unique or hard to reproduce, it is the economic basis for the franchise. We talked about Mapquest licensing its map data from Navtech, which then licensed the same data to Yahoo! Maps, MSN and Google. Google then opend the interface to maps.google.com so that other providers could add value to the maps, like Craig's list or real estate sites. (These websites that combine data streams from different web services are call "mashups.")

Service vs. product. Web 2.0 companies offer, via their websites, services to customers. They do not release their software, although it is a key asset of the company. Instead of offering yearly updates, for instance, the software inside Google or Amazon is continually improved. Since it is also constantly in use, 24x7, these companies must pay close attention to their "operational" activities (system administration, web server balancing, security, collecting usage data, etc.), not just product development and sales.

For more info on Web 2.0, read the O'Reilly article mentioned above. 

Thinking About Web 2.0 Concepts in the Context of CS73N Project Websites

 RSS. With this technology, your readers or customers no longer need to proactively decide to visit your site on a regular basis. Instead, they can subscribe to your site via an RSS feed.  Whenever you update or add new content to your site, your subscribers will receive a pointer to the updated pages. Subscribers employ free (or inexpensive), easy to use RSS readers to enable either web-based or desktop-based notifications and reading.

 The implications of offering RSS feeds on your website:

• content on the site should be updated or refreshed regularly. Many RSS readers indicate when a web site has remained the same since the last time it checked for updates. Thus, it will be more obvious to a patron if your site does not have any new content to offer.
  
• site structure should be straightforward, with clear headers and labels. When RSS subscribers click through to your site to read the new content, the structure should encourage and enable easy navigation to other parts of the site.

Collaborative filtering and capturing value from user input. Part of the special appeal of sites like Amazon and Netflix, is the ability to offer a customer the collective wisdom of all of their other customers. The companies have created huge proprietary databases from site visitor information, and then developed sophisticated software to analyze the information and use it to create unique new service offerings. Valuable visitor information comes from many sources, including

• purchases, returns, abandoned shopping carts
• free downloads
• help desk questions, FAQ helpfulness of response inquiries
  
• visitor comments, e-mailed suggestions
• registered user profiles
  
• user forums
• user-added content
• blog trackbacks
• pages visited, cookies

The overarching goal of all of this data collection is to improve the user experience, so that visitors will be inclined to come more often, stay longer, and interact more fully. Savvy Web 2.0 companies utilize visitor information for many purposes, including

• personalizing the user experience
• improving sales through targeted recommendations and/or advertising
  
• improving web sites-- adding new services, restructuring for ease of use
• demonstrating value to advertisers
• sharing information with interested third parties

The implications for student websites is to consider ways to facilitate and encourage user interactions and contributions. Users add value.