Activity 3: Perpetual Beta

Another one of O'Reilly's web 2.0 application patterns is the perpetual beta of these applications. This involves the constant updating current and adding of new features to an application, meaning that there has been a shift in the way that these programs are delivered, moving away from being a product and becoming a service.

Tumblr

Tumblr is a microblogging platform which allows its users to create their own posts or share others by reblogging them. They can also follow these blogs and have their posts appear on their dashboard, or like posts if they want to show their appreciation without the content appearing on their own blog.

Best Practices

One of the best practices that Tumblr is a great example of is that it releases early and often. This means that it is constantly adding new features to its product as well as updating and improving its existing ones. Therefore, any bugs within their system are also being found quickly and rectified. This means that users are constantly getting improvements and maintain their interest in the platform, whilst the developers are continuously getting feedback from users ensuring that they are in fact creating what the users want.

Another best practice that Tumblr is utilising is engaging its users. The staff of the website maintain major two blogs relating to this, one a general purpose one, and another dedicated to updates from their software engineers. This lets users know what updates have been made to the website, both front and backend, and how these will affect the user's experience and interactions with the website.

Comparison

Blogger is a competing blogging platform owned by Google. However, this service has its users focus on creating their own content rather than sharing the content of others, but unlike Tumblr, it does allow for commenting on posts. However, compared to Tumblr, Blogger has a lot of work to do on its perpetual beta strategy. Like Tumblr, its staff members maintain a blog on what updates are being made to the platform, but it is infrequently updated. There appears to only very rarely have new features added or updated, with 5 functionalities added or updated in the past two years, compared to Tumblr's many more.

However, Blogger does have a major focus on the best practice of making operations a core competency. Being owned by Google, its uptime can be monitored through their apps status dashboard. This releases information on whenever there has been a service disruption or outage. Through this service, Google also tells its users how they will be affected, when they expect the issue/s to be solved and if it is possible for it to be recurring. This is a great initiative by Google, as it shows their trust in their users and they make themselves publicly responsible for any problems users encounter.

Future

Tumblr has many options and paths for the future, and since it is in perpetual beta, its developers are easily able to trial new features and updates. Tumblr could improve its services by embracing making operations a core competency by releasing information about its uptime, like Blogger, and increase user trust in the platform. Another area which Tumblr could improve its operations is by engaging its users more in co-development, either by opening itself up more to user suggestions for improvements or by creating an option where users can become testers for functions that haven't been generally released yet.

What do you think of Tumblr's perpetual beta strategy? Tell me about it in the comments.

Activity 2: Rich User Experiences

Last week, I blogged about one O’Reilly’s web 2.0 patterns, data is the next Intel Inside. This week, I’ll look at another one, rich user experience. This involves delivering a PC style and level of interactivity via web browser applications.

One web application which is notable for its rich user experience is Defringe, an online art gallery. Pictures and a short description are tiled across the page and users are able to click to open an article, containing more information on and pictures of the piece.

Best Practices

Defringe exhibits many of the recommended best practices for providing a great experience for its users. Firstly, it puts usability and simplicity first. Users are instantly able to work out how to operate the website, but just as importantly, the designers found a way to make this interactivity fun. Small details, such as the way the navigation icons change or page numbers scroll as a mouse rolls over them, give the application a sense of personality, and this in turn gives the user a more enjoyable experience.

The website also puts a focus on search over structure. Content is easily sourced through typing key words into a search bar found in the navigation on each page. Although structure is provided by allowing users to sort by category or publishing date, this definitely isn’t the application’s focus as individual articles don’t disclose this information, and therefore would make it harder for a user to look for an article they had previously read based on this.

Despite having a seamless transition between various interfaces, Defringe also preserves its content’s addressability by changing the URL for each page. This is clearly a very important ability as it allows users to share their content, and hence increase their site traffic and general brand awareness.

However, one best practice which Defringe is seriously lacking in is deep, adaptive personalisation. The application doesn’t allow for users to create accounts (although due to the application’s purpose there really is no need for this functionality), nor does it make article recommendations based on past browsing activity. This could be an ability which site developers could implement later.

Comparison

In contrast, another platform based on a similar idea but with a very different execution is Instagram. This is a web and mobile application which allows users to upload pictures with a description and tags, which help others to find the post. However, this platform operates very differently to Defringe. Firstly, and most importantly, it relies on user uploads instead of Defringe’s model of having only moderators post (albeit with giving users the ability to make a post suggestion). Secondly, its web application has no focus on search. It is impossible for users to search for content using the web application unless they click on the tag they want within a post they are looking at, although third parties have filled this gap. This complete lack of functionality makes it the antithesis of Defringe, which has a huge focus on searchability. Also unlike Defringe, Instagram is actually a very good example of having deep, adaptive personalisation. Based on a user’s activity, the web application is able to recommend other users for someone to follow.

Do you find Defringe's user experience as enjoyable as I do? Let me know in the comments.

Activity 1: Data is the Next Intel Inside

All web applications rely on data, and its management is becoming a core part of a business’ strategy in an aim to get the most out of one of their vital assets. Hence, O’Reilly has made “data is the next Intel Inside” one of its web 2.0 patterns. Many companies, such as Spotify, are learning to balance having a good control over their data, whilst giving users the access and rights to their own data and the platform's, so together they can explore the Spotify's full potential.

The Platform

Spotify is a web, desktop, and mobile application which allows users to stream millions of music tracks; follow artists, playlists, radios, and friends, as well as share libraries and playlists with friends. Based on a user’s activities, the application is able to suggest other music the user may enjoy listening to. The company is meeting one of O’Reilly’s best practices by designed their data for reuse. They have released an API, allowing outside developers to fetch Spotify’s data relating to artists, albums, and single tracks and users’ personal playlists and libraries (with their permission).

The Competition

Compared to other common music streaming platforms, Spotify is a leader in API offerings. YouTube is one of their main competitors, however it is focused on playing a single video (or much less often, a playlist), instead of Spotify’s more specialised constant streaming of audio. Both platforms allow the retrieval and streaming of tracks/videos or playlists, however Spotify does fall behind when compared to YouTube’s Analytics API, which allows developers to access viewing statistics, viewer demographics, and popularity ratings.

Spotify’s other major competitor is iTunes, which also has released an API. However, the iTunes API has a lot fewer capabilities. Like Spotify, it allows for developers to search the store and display a song or album’s information, but unlike Spotify, it only allows for a small preview of a song to be played. Most crippling though, is that Apple (the owners of iTunes) don’t allow the API to be used for any purpose except to promote the store (even going as far as requiring an iTunes icon displayed wherever it’s used), and explicitly banning entertainment usage. This means that Apple is missing out on many opportunities for its data to be used and exposed to developers with ideas, allowing for Spotify to swoop in and gain a huge market share.

The Future

One best practice Spotify must work on is allowing users to control their own data. For example, there is no way for a playlist to be exported to another application such as iTunes or Windows Media Player, even if it consists entirely of the user’s own locally stored music files. Another possible future direction for the platform is focusing on another best practice, enhancing their core data, by creating opportunities for users to make more explicit inputs, such as ratings and reviews.

What are your opinions on Spotify's use of its data? Let me know in the comments.