Open Air

Being successful at Web 2.0

October 31st, 2008 Posted by: l.dewis

At last week’s Web 2.0 Expo I went to see Dion Hinchcliffe talk about Building successful next generation Web 2.0 applications.

Here are the most pertinent points for the OU as it considers building SocialLearn.

Design is what makes the difference between a leading and an average site. Growth is only limited by design of your application. 95% of sites have to drive traffic constantly, and when the product stops being promoted it stops growing. YouTube and FaceBook both grew to lead positions within 18 months without any major promotion.

The success is down to the user providing the value and having control and the site responding quicker than the competitor to the needs of the user. One example is XM radio who gave control of the station to their listeners – what would happen if the OU gave control to the students? Can these models apply to education? Is a course list like a playlist? What is the equivalent of a DJ shout out in a university? Other examples include Amazon’s mturk.com – a marketplace for work, where the worker selects a task and works when they want to, or castingwords.com transcription services. People are embedded in product, people are what make the product.

If we accept that the user creates the value in Web 2.0, it follows that your user differentiates your product from any other. If good features get quickly and easily copied (like status updates) your USP may be in your users and how well you respond to their needs. How can students enrich what a university offers? How can we respond better to students needs? Can we get over the idea that we (the institution/teacher) always know best? Users of a product often know the product better than the people who work for the company who make it, and will use it in unexpected ways. In education a teacher will be intimately connected with the ‘product’ (course) but will not experience it in the same way their students will.

Dion said that competitive advantage is in having the best data and releasing new features/ capability every few days, continually responding to the market place quicker than your competitors. We have to see SocialLearn as a product with customers, rather than an OU supported project if we’re to work in the ways necessary to succeed.

The value in collective intelligence is in user contributions, building up a database of intentions (searches, clicks, tags, recommendations), sharing that data to improve user experience (prior similar uses, accumulated experience), and creating viral feedback loops.

An architecture of participation fundamentally underpins Web 2.0 sites like digg. If the value is in user contributions and 1-10% of users contribute, how many users do you need to create a valuable site? Are you paying lip service to Web 2.0 if your site could exist without your users? There is a danger in designing a site that says we are the best because of what we know, rather than saying we are the best because we facilitate our users to know what they want to know better. Transparency is essential to showing everyone who did what and the more you know about the user the more you can place a personal, educational or commercial value on that person’s action. I’d like to see SocialLearn help identify the ‘connectors’ in educational networks that enable information flow between niche subject communities.

It’s all about intentions, gestures, personal data, shared data, and a secret formula!

Thinking about the data set that best captures educational experiences is vital to the success of any project like SocialLearn. If potent datasets are behind all leading sites and the value is in information, what data can we capture from learning and feedback to users to help them learn more efficiently? The race is on to own the major classes of data online and given that use adds the value, access to that data will be free. This means that people will only use the best set of data online. They have no reason to use second best. Social bookmarking is still a class without a strong owner, as is identity, public calendaring of events, and parental control data. Search will act as a lens on user data. Can SociaLearn act as a lens on learning data?

And how can it utilise other API’s that give access to data. Comparing data sources can be really useful for students. One example is Flash earth – the user chooses to view the same map from different perspectives by choosing data from various map data providers. Destinations sites are going to fail, and revenue comes from API use. 90% of Twitter use comes via the API. Amazon have 55,000 partners using their ecommerce API’s live. They turned a web service into a platform and 300K businesses build on top of what they’ve produced. Examples include simple storage service (s3) and elastic compute cloud.

Other value comes from SDK’s. What is the OU version of a Google web toolkit?

Some sound advice from Dion included:

• Think about cost effective scalability – 24/365 operations
• Maintain control over unique hard-to-recreate information that gets richer the more people use it
• Trust users as co creators
• Harness collective intelligence
• Go for lightweight business models – customer self service, long tail, turn apps into platforms, encourage unintended uses
• Either meter use of your api like electricity meter or give it away for free if use will contribute to your sales.
• You need skills in assembly over development – mashup’s seamlessly bring content from several places into one place making innovation in assembly more important than ingenuity in coding.

He also covered some of the issues involved in choosing to use 3rd party services. Figure out what we do better than anyone else, focus on that and use 3rd party services for anything else.

Entry Filed under: Web 2.0