ChatRoulette - Fail Fast, Scale Fast

Watch This Blog Above

There wasn't too many Internet start-ups in the years following the 2000 dot com bust. However, ten years later the story is very different and it has nothing to do with venture capitalists re-entering the market rather lightweight, scalable business models that have been made possible through cheaper hardware, free open source code, free marketing and distribution and powerful programming languages reducing the need for large development teams (Graham, 2008).

Competitiveness in the global online market relies on developers delivering products before their competitors and when they arrive they need to be cheap. Speed is of the essence as Van Grove (2010) puts it "the company with its name in lights is the company that most often will prevail". She was speaking about a recent start up ChatRoulette which is a service where users have video conversations with random users around the world. ChatRoulette is a great example of a business which experienced growth at a viral level a scaled appropriately.

Chatroulette Screen Shot

How a company receives revenue is directly related to how much the consumer pays for the product. Using the example of ChatRoulette the service is provided for free and revenue is generated through advertising. Costs a kept low because the bandwidth of the video is not ran through the server (CamChat, 2010). ChatRoulette's founder is quoted saying he turns a profit through such advertisement. Watson(2010) has suggested it is important for such companies to diversify possible revenue streams because the advertising market cannot alone be relied on.

Viral marketing is a cost effective method of promoting a new product and it doesn't happen by accident. Web pages such as YouTube or new.com make it easy for users to share their content with their friends using buttons such as email or share this article. Considering our example of ChatRoulette Andrey Ternovskiy (founder) used internet forums to publicize his new service (Bidder, 2010). Word of mouth caused his service to go viral and move from 500 users per day to 1.5 million. To cater for the sudden growth in demand Ternovskiy relied on outsourcing.

Businesses can access server storage and bandwidth as a service meaning they only pay for what they use (see one services pricing below). Chatroulette used a similar cloud service which enabled them to quickly scale with demand when the service went viral. Ternovskiy also outsources development. He is currently employing four programmers from around the world who are working on improving the service. Outsourcing provides companies such as Chatroulette cost effective flexibility and allows them to focus on their core activities (Bucki, 2010).

Linux Server Size (RAM / Disk)	Hourly Cost *	Monthly Cost
256 MB / 10 GB	1.5¢	$10.95
512 MB / 20 GB	3¢	$21.90
1024 MB / 40 GB	6¢	$43.80
2048 MB / 80 GB	12¢	$87.60
4096 MB / 160 GB	24¢	$175.20
8192 MB / 320 GB	48¢	$350.40
15872 MB / 620 GB	96¢	$700.80

(RackSpaceCloud, 2010)

By utilizing re-usable code, viral marketing, multiple revenue sources and outsourcing hardware services can experience fast growth and make quick returns on relatively (to traditional models) small investments (O'Reilly, 2005). Chatroulette took 17 year old Ternovskiy 2 days and 2 nights to create and is now worth an estimated 10 million euros (Bidder, 2010). He focused on the core functionality, released quickly and without the need for venture capitalist . This new business model is known as lightweight and scalable and is profitable (Zawodny, 2004).

References

• O'Reilly, T. (2005). What is Web 2.0.
• Bucki, J. (2010). Top 7 Outsourcing Advantages.
• Zawodny, J. (2004) Lightweight Business Models at Web 2.0.
• Van Grove, J. (2010). Why Chatroulette is More Than Just Penises.
• CamChat. (2010). CamChat Server Requirements.
• Kondakov, Y. & Bidder, B. (2010) . Mom, Dad, the Site Is Expanding.
• Graham, P. (2008). Could VC be a Casualty of the Recession?.
• Rachspace Us Inc. (2010). How we Price Cloud Servers.
• Watson, J. (2010).Week 10 Netcast: Lightweight and Scalable Business Models.

The Long Tail of Steam

New! Watch this blog on YouTube (see above)

Traditionally (before e-commerce), for a product to be successful it needed to have high-volume appeal. However, due to cost reduction and better targeted marketing (possible due to the Internet), money can be made selling to niche markets. Collectively niches can generate sales numbers similar to or better than traditional markets (Davis, 2005).

Steam is an online digital entertainment company that currently offers 1100 games for purchase, download and play from any computer (Steam, 2010). Compared to department stores that rarely offer more than 100 titles (usually the more popular ones), Steam’s range is simply larger and caters better to specific interests rather than current trends. Such business models are what Anderson referred to when he coined the term leveraging the long tail (Anderson, n.d.).

Traditionally retailers only targeted the ‘head of a market’. However, collectively, products in the tail have the potential for similar or greater sales (Anderson, n.d.).

Companies such as Steam are able to cater to these niche markets by using new technologies to reduce costs (Wikipedia, 2010). Users enter, edit and update their personal information negating the need for some customer service employees. Steam users enter their own details from payment information through to their billing address (Steam, 2010). Self service extends to support, using the Steam forums and knowledge base gamers can ask their peers technical questions. This reduces the amount of money spent on support staff.

Products purchased from Steam are delivered virtually over the Internet in a digital form (Steam, 2010). In contrast to traditional business models, Steam does not pay for products until they have been sold to a customer. Furthermore, like other companies that trade in digital content there are few costs associated with a physical retail front or delivery.

Where the cost of inventory storage and distribution is low it becomes profitable to sell relatively unpopular products (Wikipedia, 2010). As a result vendors no longer need to put consumers in a one size fits all container (Anderson, n.d.). However, as the tail lengthens, consumers may find it hard to navigate to products of interest. Sites such as Steam use algorithmic data management to match supply and demand (O’Reilly, 2005). Based off pages a user visits, Steam recommends other games they might be interested in.

In summary, as Anderson (n.d.) suggests, People gravitate towards niches that are narrow interests (and everyone has them). Trends show that through using new technologies to reduce costs collectively, niche markets are as economically viable as their traditional high appeal counterparts. Businesses can now be built on hitting the growing niche market (Holter, 2006).

As for the future of Steam, I would like to see them follow the trend of the Apple App store and allow more independent developers to submit games for sale (this would lengthen their tail too).

References

• Anderson, C. (n.d.). About Chris Anderson.
• Davis, H. (2005). The Long Tail.
• Holter, E. (2006). Examples of the Long Tail Effect.
• O’Reilly, T. (2005). What is Web 2.0?.
• Steam. (2010). Steam Store.
• Wikipedia. (2010). E-Commerce.
• Wikipedia. (2010). The Long Tail.

Mindmeister: The Perpetual Beta without the Beta Tag

“There's no final version. Nothing is static, everything is changing.”(Chitu, 2007). This is how Google defines the idea of perpetual beta, software as a part of an indefinite and continuous development cycle (Wikipedia, 2010). This is the new way of thinking about software development and one that gives young companies an advantage because they are not in the habit of the traditional software release cycle.

Mindmeister is a collaborative mind mapping tool that demonstrates the Web 2.0 principle of ‘Perpetual Beta’ and has seen success through these innovations. Meisterlabs has made operations a core competency. Their security page boasts 99.9% uptime and the highest data center standards (Mindmeister, 2010) and their product is updated many times a month. Such measures are essential for web 2.0 applications that are offered as services because they won’t perform well without regular maintenance (O’Reilly, 2005).

Mindmeister user numbers.

Meisterlabs subscribes to the benefits of releasing their software early and often. Through the change log it is clear that their service is updated multiple times per month. The improvements made however, are usually small and not always noticeable to users.

Improvements are able to be regularly rolled out through utilizing light weight, flexible and cross-platform programming languages. O’Reilly (2005) refers to such examples as php, AJAX or ruby on rails as prime examples of code which can be written quickly to enable responsiveness. Mindmeister uses HTML, CSS and AJAX client side to allow their application to be rendered by most browsers. Ruby on wheels is used server side (Hollauf, 2010).

Mindmeister uses beta testers who post reports on bugs they identify. Mindmeister aims to be responsive to these reports. O’Rielly(2005) identifies the use of shadow applications which can collect data on how an application being used. These should be planned and implemented with the main application. Based on the data (see below) Mindmeister is able to provide it can be assumed they use shadow applications. Other sites such as Google and Amazon have been know to engage a small percentage of their users in such tests before rolling out products. Testing is used trial new business strategies and application functionality.

Web 2.0 means consumers are now purchasing services instead of artifacts. These services need to be maintained to respond to user requirements and through leaving an application in perpetual beta products see a faster time to market, reduced risk (less up front cost), a closer relationship to customers and the ability to quickly and effectively respond to real time data (O’Rielly, 2005).

Question for thought:

Will we see the Beta label return to products when large versions are released?

References:

Chitu, A. (2007). Perpetual-Beta.

Hollauf, M. (2008). Mindmeister: Development Cast Study.

Meisterlabs. (2010). Change Log.

Mesiterlabs. (2010). About Security.

Meisterlabs. (2010). Mindmesiter Launches Private Beta Test.

O’Reilly, T. (2005). What is Web 2.0?.

O’Reilly, T. (2005). What is Web 2.0: End of the software release cycle.

Wikipedia. (2010). Perpetual Beta.

Location, Location, Web 2.0 – Foursquare is Built From Many Devices

Foursquare is a location based social networking game. Users ‘check in’ at locations announcing they have been there in competition to become mayor. People contribute through submitting their thoughts, locations, pictures, information about their current location and friends (Wikipedia, 2010).

Utilizing features only available on cell phones has heavily contributed to Foursquare’s functionality and success. Photos are taken using cameras, locations recorded using cell location services and the application relies on a phone’s connectivity (Foursquare, 2010). Combined, these features have created an application that couldn’t exist solely on a desktop computer or phone.

In the company of desktop computers, Foursquare can be used on iPhones, Android phones, Blackberrys, Windows mobiles and systems running webOS (Foursquare, 2010). An application is currently being developed for Nokia handsets (Guim, 2010). The way each platform is used to its strength and data is shared between devices and services, makes Foursquare a great example of software above the level of a device.

O’Reilly (2007) uses the example of iPhones being managed by iTunes to show how software can work in tandem to ensure a better user experience. Foursquare follows a similar model allowing users to manage their profiles online. This means users aren’t required to type or read great amounts of information into/from their phones. The mobile applications are then left to perform the task they are great at, being mobile and more so, immediately recording users’ thoughts or pictures in the context of location, harvesting rich forms of media.

Foursquare intends to generate revenue through advertising. Businesses can advertise directly, targeting people who are regularly nearby their businesses in the form of couponing (Carlson, 2009). Data generated by Foursquare has been used to create mash-up applications.

Fourwhere, a desktop web application, uses collected data to display comments made through Foursquare on a map. This is a great example of how Foursquare extends the usability and functionality of their data using the context of location (Sysomos, 2010).

Notably, Foursquare encourages usage by allowing users to import friends from Facebook and Twitter. These services are also used to broadcast a users location.

In summary, Foursquare utilizes the strengths of many devices, including a broad range of phones and desktop computers, to harvest rich media in the context of location. Thinking forward we may soon see such applications incorporating the accelerometer and future phone features as they are released and patterns indicate this trend may lead to richer data being harvested without an increase in conscious human interaction.

Questions for thought:

What other data types could be harvested without the need for human input?

References:

Butler, E.(n.d.).FoursquareX

Carlson, N.(2009).How Will Foursquare Make Money?

Foursquare.(2010).Welcome to Foursquare

Guim, M.(2010).Foursquare App for Nokia N900

O’Reilly, T.(2007).Software Above the Level of a Single Device.

Sysomos Inc.(2010).FourWhere

Wikipedia.(2010).Foursquare (Service)

Word processing accessible by everyone

Google Docs is an online productivity suite offering a word processor, spreadsheet and presentation application (Google, 2010). Despite being a web application, its functionality and user interface places it in direct competition with traditional desktop applications such as Microsoft Word and iWork.

Docs boasts many advantages over its desktop counterparts. Documents:

• are easily shared, instantly updated and can be accessed by multiple people simultaneously,
• offer added functionality through application programming interfaces such as Google Lookup and
• are backed up on many servers to ensure they are not lost.

To encourage collaboration and ease of access, all documents are searchable and uniquely addressable through a URL. Furthermore, Google Docs ensures files can be accessed from any location with an internet connection, without the need to install an application (Strickland, n.d.).

Web browsers are responsible for rendering Docs’ user interface (UI) and a majority of the heavy processing is done server-side. Most popular web browsers are supported (see below) and when cross-referenced with browser usage statistics we find that 88.3% of the world’s browsers can use Docs. HTML and Javascript is used to render the UI client-side. AJAX script creates regular links to the server instantly updating the document when it is edited (Strickland, n.d.).

Browsers Supported by Docs

	Windows XP / NT	Windows Vista	Linux (Ubuntu)	MAC OSX 10.4 +	Mac OSX 10.3.9 and below*
Google Chrome	X	X
Safari 3				X	X
Safari 4				X	X
Internet Explorer 7	X	X
Internet Explorer 8	X	X
Firefox 3.0	X	X	X	X

(Google, 2010)

Browser Usage

– These statistics are approximate. Statistics vary source to source.

2010	IE8	IE7	IE6	Firefox	Chrome	Safari	Opera
February	14.7%	11.0%	9.6%	46.5%	11.6%	3.8%	2.1%

(W3, 2010)

Docs is subject to limitations common amongst rich internet applications (RIAs). Browser, broadband speed and reliability limitations means Docs can be slower to access data and respond to user input when compared to desktop applications. File storage is limited by the allowance offered by Google (currently 1gb free). Furthermore, security concerns are often raised surrounding documents being stored online. Finally, Docs functionality is limited compared to the likes of MS Word (Strickland, n.d.).

ThinkFree is a word processer, similar to Docs. However, before using the RIA users download a java applet, this allows the application to offer greater functionality. This may be an avenue Google may need to take in the future to compete with the functionality of desktop applications (Gottipati, 2007).

Google docs and ThinkFree represent only a couple of RIAs currently on the market, with many more available or in development. Compliance with HTML5 standards will require browsers to perform tasks traditionally performed by operating systems and RIAs may replace a majority of desktop applications. This shift will see these services (applications) offer consistent interfaces and functionality, to everyone, anywhere, at anytime without the need for high powered processors or installation.

With the global movement towards RIAs is there any applications that won’t work as an RIA?

References:

Google. (2010). Welcome to Google Docs.

Retrieved 25 March, 2010 from https://www.docs.google.com/

Google. (2010). System Requirements.

Retrieved 25 March, 2010 from http://docs.google.com/support/bin/answer.py?hl=en&answer=37560

Gottipati, H. (2007). Is Java more effiecient than AJAX for advanced web apps.

Retrieved 25 March, 2010 from

http://www.oreillynet.com/onjava/blog/2007/01/is_java_more_efficient_than_aj.html

Strickland. (n.d.). How Google Docs Works.

Retrieved 25 March, 2010 from http://communication.howstuffworks.com/google-docs.htm

W3 Schools. (2010). Browser Statistics.

Retrieved 25 March, 2010 from http://www.w3schools.com/browsers/browsers_stats.asp

Wikipedia. (2010). Google Docs.

Retrieved 25 March, 2010 from http://en.wikipedia.org/wiki/Google_docs