This week we are talking about lightweight models of Web 2.0 applications that have cost-effective scalability.
There are a few key elements that have encouraged the design of these lightweight models. One is the advent of open source programming to reduce the difficulty in producing web 2.0 applications. Linux , mySQL and PHP are all free of charge examples. Also, traditional broadcasting was expensive whereas podcasting is basically free of charge. These elements ensure a cost-effective site is designed.
Previously, large companies would have to invest in large time-consuming projects to buy servers and software whereas a Web 2.0 application can start small and has the ability to build to a much larger application overnight if need be. This business model and technology scalability needs to be inherent from the start.
Ensuring the application has high responsiveness is a another key element. To maintain this, all aspects of the application should be outsourced. This could include storage, web hosting, coding and maybe even customer service. Outsourcing ensures the most cost-effective option is taken while enabling quick responses to increases or decreases in site use.
Using the ecosystem of developers that are available online ensures flexibility. Site founders request help from developers and if the idea fails only hundreds of dollars are lost and not millions like previous eras. If they are not sure, this flexibility allow them to choose another developer.
An inflection point occurs when suddenly (often overnight) there is an exponential growth in the number of users and page views of a site. This is clearly demonstrated on the graph here. The number of users grows slowly over time but it hits a point where the number of users becomes near vertical and reaches new heights.
It is suggested that if you can get to this point first, you should be safe from competition as it becomes difficult for the competition to unseat you when you have that many users. But, if you hit the inflection point without a revenue model then you are not viable. It is a misconception to think that no revenue model is a viable approach (Watson, 2012).
With so many users trying to access the site, there becomes an instant need for more servers. This all costs money. Therefore, there is a need to make revenue to ensure the site remains accessible. To ensure a highly quality revenue model, there needs to be multiple revenue avenues. Here are a couple of examples:
Syndicated advertising uses many components that have already been built by others. This ecosystem can then earn advertising revenue by providing information for other companies. Adsense is an example. This is “leveraging the long tail” as a new site’s products will likely be a niche market and the advertisements can be contextually relevant to those products.
Charging a subscription in a tiered service model will get revenue, unfortunately the ‘cost’ simply puts people off. To avoid this, the first tier can be a free service and then a freemium charge will get extra benefits.
Tumblr was founded in 2007 by David Karp, a high-school dropout who was then 20 years old and who launched the site to simplify the blogging process.
The name stems from tumblelog, which is a short-form mixed-media blog.It is a simple, customisable micro-blogging platform that allows users to post images and video or audio files in addition to text. Users say if it’s too big to tweet and too short for a substantial blog post, they post it on their Tumblr blog. Members of the site can follow other Tumblr pages that can be viewed on their dashboard, and can “reblog” or “heart” other user’s posts.
“Tumblr was an accidental social network”
founder and CEO David Karp
It started with two employees in a tiny Manhattan office. In its early days, the service didn’t have the “steep social network growth” because it was about a core community of creators. The company originally set out to build novel tools that offered an escape from the restrictive templates of Facebook and Twitter (Gannes, 2012).
Within two weeks of its launch, Tumblr had 75,000 registered users. It continued to grow in 2010, adding 10 times that many every month, and took about $10 million in venture funding (Ingram). Between 2010 and 2011, Tumblr grew an impressive 166%, reaching 10.7 million visitors in May 2011. At that stage, Tumblr was clearly experiencing a viral adoption curve and reached the point of critical mass threshold that propelled it to widespread adoption (Lipsman, 2011). Today, as you can see below there are 18.5 million visitors (quantcast, 2012).
The inflection point (as discussed above) for Tumblr looks to be about half way through 2009 where there is a curve up in the number of users.
Continually growing at over 30% a month has not been without challenges. At peak time there are 40,000 requests per second and 3TB of new data to store a day, all running on over 1000 servers (Hoff, 2012). This is expensive!
Tumblr was originally funded by Karp’s earnings as a software consultant for online parenting site UrbanBaby (Wikipedia, 2012). Karp raised $4.5 million from Union Square Ventures, Spark Capital (key investors for Twitter also) and Betaworks in December 2009, which gave Tumblr plenty of money to stay focused on attracting users. In 2010 Karp suggested revenue generation by customizing its platform for big media companies and by charging regular users who wanted extra services, such as more visibility for their pages but this was not a substantial revenue raiser (Ricadela, MacMillan, Hesseldahl, Kharif, 2009). At the end of 2010, they raised between $25 million and $30 million in additional financing and the business was valued at an estimated $135 million (Primack, 2010).
Tumblr is now thought to be worth $800 million (Ante, 2011). On the other hand, its costs are growing rapidly and on May 2nd a brand new revenue model was introduced to cover these costs. The model provides a way for brands to connect with consumers. In the new scheme, Tumblr users can pay to have their blog posts appear on the site’s main page providing an opportunity to push out brand’s messages without the stigma of formal ads (Danna, 2012).
Low barriers to entry into designing Web 2.0 applications (costs of next to nothing) can be a double edge sword. Someone could see Tumblr’s simple idea and then create an equivalent site that is quicker and better. So many people are out to make a quick buck, but only the strongest will survive – those that have a good revenue model. The model needs to be diverse to ensure long-term stability.
When it comes to scalability, there needs to be a plan to ensure that all the information above is available to all users at anytime. In the last year, Tumblr has hired around 90 people (up from fewer than 20 in 2011). To accommodate the more than 16 billion monthly page views, the site has also had to increase its server capacity, its support staff, and its office size. It now occupies two floors in a loft building in Manhattan’s tony Gramercy neighborhood (Danna, 2012).
Traditionally, Web 1.0 app designers aimed to get big fast but now with Web 2.0 – small is the new big. Wikipedia is a good example with only 22 employees and is the 4th largest site on the internet (Watson, 2012). As you can see, Tumblr is another great example from the statistics above. This makes for a very cost-effective business.
Tumblr has also been in breach of copyright rules as the hosting site, if there is a complaint Tumblr has to pull down the content immediately. Tumblr suggests being fair and using proper credit and attribution if sharing content that wasn’t originated by the users themselves.
Outsourcing leaves a company open to copying from outsourcing company where they then become the competition. They see what they call a gap in service and begin providing the service themselves.
The future looks bright for Tumlbr, it is predicted that there will be more users than the full blog sites like WordPress shortly. Users like simple and easy tools to communicate with their followers and Tumblr gives them this opportunity.
Tumblr management needs to allow the staff to the lead – they are normally the ones that circumnavigate a process that is difficult. Looking at ways to make things very easy for the user is the key, this is referred to as Social Technology.
Tumblr has become so popular as it uses so many of O’Reilly’s design patterns. It leverages the long tail by allowing users with niche content to posts their goods. It uses collective intelligence as the users are making the content on the site. It is continually rolling out new ideas so is said to be in perpetual beta. With the introduction of mobile phone apps it is software about the level of a single device. With a simple microblogging design, it has become a rich user interface.
Ante, S. (2011) Tumblr Valued at $800 Million. Retrieved on 12th May 2012.
Danna, R. (2012) Tumblr’s Growing Pains: Can the Blogging Site Cash In? Retrieved 12th May 2012.
Gannes, L. (2012) Tumblr’s Inflection Point Came When Curators Joined Creators. Retrieved on 10th May 2012.
Hoff, T. (2012) Tumblr Architecture – 15 Billion Page Views A Month And Harder To Scale Than Twitter. Retrieved 13th May 2012.
Ingram, M. (2010) Google VC, Tumblr CEO Among the Top Innovators Under 35. Retrieved on 10th May 2012
Lipsman, A. (2011) The Network Effect: Facebook, Linkedin, Twitter & Tumblr Reach New Heights in May. Retrieved on 11th May 2012.
Quantcast (2012) Tumblr Blog Network. Retrieved on 12th May 2012.
Primack, D. (2010) Tumblr dives into a boatload of money. Retrieved on 12th May 2012
Ricadela, A., MacMillan, D., Hesseldahl, A., Kharif, O. (2009) Best Young Tech Entrepreneurs 2009. Retrieved 12th May 2012.
Watson, J. (2012) Lecture 10 Notes for QUT INN347 Web 2.0 applications.
Wikipedia (2010) Tumblr. Retrieved 12th May 2012