New Market Disruptive Innovation for Web 2.0

July 2nd, 2008 seph250

New Market Strategies for Disruption

New market strategies for disruption are different than low-end disruptive strategies.  Instead of simply entering the market with a lower cost, lower performance product, new market strategies involve “coming off the backplane” into a new market.
An example of a new market disruptive innovation is Sony’s portable transistor radio.  When originally released, the transistor radios sound quality was completely inferior to that of existing radios on the market.  However, existing radios were so large they were often built into furniture and thus not remotely portable, and prohibitively expensive for most customers.  When Sony’s low cost portable music player came on the market, it was not really a direct competitor to the existing radio businesses.  If someone was in the market for a high end radio, they would balk at the poor sound quality of the transistor radio.  Likewise, most customers interested in the transistor radio could never have afforded an expensive home radio.  By positioning the product for a new market, Sony was (according to  Christensen) “competing against non-consumption”.  Customers were either going to get a Sony radio, or not get anything at all.

New Market Strategies to Disrupt a Web-Service

In most of the examples Christensen describes, new market disruptive strategies seem similar to low-end disruptive strategies because price is the factor that separates different markets.  In free consumer internet services, markets are not defined by pricing categories because the prices are already zero, a price that anyone can afford.  However, there are other factors that separate markets in consumer internet and thus still ways to position a service such that it “competes against non-consumption”.

New Market Disruption

Foreign Language Applications

One proven strategy is to offer a consumer internet service in a certain foreign language.  Many consumer internet sites target the US market and ignore the rest of the world.  If you don’t speak English, you can’t use these services.

Xiaonei Disrupting Facebook

An example of a company employing this strategy is Xiaonei, dubbed by many as the “Chinese Facebook Clone”.  In April 2008, Xiaonei reportedly raised $430 million in venture capital, and had built a community of over 8.8 million active users.  As in all disruptive innovations, it currently seems that the two social networks do not compete with each other.  As globalization continues it’s possible these companies will become direct competitors.  If Xiaonei is able to focus on the unique needs of their Chinese customer base and to develop real innovations in their own right rather than just mimicking what Facebook does, it may be able to build enough momentum to compete with and even disrupt Facebook in the English market.

Baidu Disrupting Google

Similarly, Baidu is a search engine that targets the Chinese market.  The interface, while resembling Google’s search page, is in Chinese, but the differences go much deeper.  For a search engine to adequately address the Chinese market, it needs to be able to index and search through pages that use Chinese characters.  This provides unique challenges that companies focused on the US market are not well positioned to address.  In 2006, Baidu was “the first choice of 62 percent of Chinese users, up 15 points over 2005”, according to a study released in September by the China Internet Network Information Center.  As globalization continues, Baidu may one day expand into offering a US search engine that can disrupt Google at home in the US.

Provide Mobile Access

Another new-market disruptive strategy is to initially focus on mobile.  Around the world, many people have access to a mobile phone, but no access to a computer.  Mobile services are generally lower performing products than their personal computer browser counterparts limited by the smaller screen, user interface and slower data rates.  Some mobile services work by simply sending text messages to the service, which responds with a text message, others work in the mobile phones browser, and others work as a Java application installed on the phone.  Mobile services with value propositions similar to desktop browser internet applications could provide a foothold to disrupt the leading internet application.  Some companies that seem to be employing this strategy include 4Info and Babajob.

4Info Disrupting Google

4Info was one of the early local/mobile search companies that functioned via text messages.  In order to submit a search a user only has to send a text message to short code 44636 (4info), of what they want to search.  Sending the message “Starbucks Boston” will return the address and phone numbers of Starbucks locations in Boston.  Sending “SWA 436” will return the details of Southwest Airlines flight 436.  By using text messages as the interface to the service, 4info is available to anyone with a text message capable mobile phone.  People without a broadband internet connection on their PC are able to use the service.  Unfortunately for 4Info the leading search engines have recognized this service as a potential disruption, and launched similar services, making it seem unlikely that 4Info will be able to disrupt Google or Yahoo.

Babajob Disrupting Monster

Another startup that is using mobile access as a new-market disruptive strategy is the India based startup Babajob.com.  Babajob is a service to connect employees with employers.  The business takes advantage of the fact that although most workers in India do not have personal computers with internet access, they have text message capable mobile phones.  Once signed up, job seekers get text message notifications of job opportunities they might be interested in, and can respond with their availability.  As the Indian job market grows and more Indians begin to use personal computers with internet connections, Babajob may be able to move up market and find itself in a position as the Monster.com of India, and a disruptor of the US market leader.

Posted in Internet, Startups | No Comments »

Low-End Disruptive Innovation in Web 2.0

June 30th, 2008 seph250

Introduction to the Innovators Dilemma

In the Innovators Dilemma and the Innovators Solution, Clayton Christensen creates a framework explaining how established companies, with large amounts of capital and loyal customer bases, are disrupted and eventually replaced by smaller companies with little capital and no customer base.  This model of disruption introduces several insightful concepts, but fundamentally says that by offering products with less performance, but at a significant price discount, new companies have the opportunity to take initial market share and then move up market and disrupt the incumbents.

With consumer internet web services these rules may no longer apply.  YouTube, Facebook, Skype, Kayak, and Pandora are free services for consumers.  If it’s impossible to undercut these companies with lower prices, is it impossible to disrupt them?  This paper will apply Christensen’s framework to web services that are free to consumers and examine possible strategies for disrupting market leaders.

Consumer Internet – Services for Free

In 2008, Chris Anderson, the editor of Wired published an article titled “Free! Why $0.00 Is the Future of Business”.  Anderson discusses how, for web services, the swiftly decreasing cost of bandwidth and data storage combined with the ability to simultaneously serve millions of customers at once are driving prices to zero.  Some of the free services that will be examined include Google, Yahoo, Hotmail, Flickr, Facebook, MySpace, Twitter, Wordpress, Pandora, Monster, and Skype.

But hold on, if these companies don’t charge for their products, how do they make money?  Two of the most popular business model’s applied to these types of businesses (as coined by Chris Anderson), are Freemium, and Advertising.

Freemium

In the freemium business model, the basic service is free, however, a premium version of the product costs money.  An example of this is Skype.  Calls between two computers running Skype are free, however, to call from a computer to a landline or mobile phone the user must pay a charge.  By offering the basic version of Skype for free, Skype has been able to garner a large number of users, develop initial relationships of trust, and later up-sell customers to the premium product.  Low bandwidth and low data storage costs mean that even if only a small percentage of people pay for the premium version of a freemium service a web-service can still be profitable.  Other web services that use this business model include Flickr, which offers increased storage space for $25/year and Pandora which lets customers stream music on their mobile phone for $36/year.  Both of these companies provide a free service as well which the majority of customers rely on.

Advertising

Online Advertising Business Model

With the advertising business model, a website offers its service for free to consumers, but charges a third party the right to advertise on the website.  As illustrated, the web service attracts the attention of the consumer, and effectively sells a portion of this attention to advertisers. Advertising is the most popular strategy for monetizing a web site and can easily be used in combination with other monetization strategies. Examples of web services using this business model are Facebook, Google, and YouTube.

Disruptive Innovation

Disruptive innovation is described by Clayton Christensen as the “Innovator’s Solution” because it seems to provide a unique opportunity for new companies to replace market leaders.  Other types of innovation include sustaining innovation and revolutionary innovation.  Sustaining innovations are the incremental improvements in products over time.  Sustaining innovations are generally predictable and it is rare for a startup to disrupt a market leader with a sustaining innovation.  Revolutionary innovations are innovations that suddenly offer huge improvements in performance.  Today, these types of technology improvements seem to require huge investments in research and are difficult for startups to achieve, although in the situations where this has occurred, great companies have been created very quickly.

Disruptive innovation is a highly desirable strategy for entrepreneurs to follow because it allows a startup to compete with established players in a way that tends to cause the leader to abandon the less profitable customers and move up market.  Disruptive innovation can be divided into two categories, low-end disruptive innovation and new-market disruptive innovation.

Low-End Disruptive Innovation

Low End Disruptive Innovation

Low-end disruptive innovation is shown in.  This type of disruption relies on the pattern that sustaining innovations improve a products performance faster than consumers can utilize this performance improvement.
An example of a low-end disruption is flash drives disrupting rotational hard discs.  Early on, consumers were underserved by the storage capacity available and placed a high value on the larger hard drives.  As the size of operating systems and software applications increased, consumers demand for storage capacity (which varied depending on the consumer) generally continued to increase.  However, with sustaining innovations, rotational hard disc capacity increased at a faster rate than most consumers could utilize.  Today, with loads of digital pictures, movies, music, and applications, the typical consumer can utilize a very large hard disk.  Even with all these demands, however, we are finding that we don’t need the super large capacity hard drives that are available.  Since we don’t need them, we refuse to pay a premium for them, and instead opt for the midsized hard drives that are available.  Lack of differentiation forces the price of these discs down, and they become less profitable.

Meanwhile, another technology for data storage, flash drives has been being developed.  Initially, flash drives had much lower capacities than rotational hard disks, however through sustaining innovations, they also increased their performance.  As rotational hard disks exceeded typical user’s needs, flash drives began to meet consumers minimal storage requirements.  Thus, flash drive storage is beginning to disrupt rotational hard disc storage.  The recently launched MacBook Air is one of the first mainstream laptops to be offered with a flash drive instead of a rotational hard disc.  As Flash prices continue to decrease, we can expect rotational hard disks to become an obsolete technology.

Low-End Strategies to Disrupt a Free Web Service

Traditional low-end strategies require undercutting the competition on price.  With prices of consumer internet services at zero, it seems this is not possible.  In this section, I will explore two variations of low-end strategies that a startup can consider: less than free and low-time disruption.

Less than Free

One approach is to actually pay your customers to use your service.  This approach was attempted by All Advantage in 2000 and is currently being employed by several video sites.

AllAdvantage.com

In 1999, AllAdvantage offered to pay customers to use their service.  AllAdvantage’s slogan “Get paid to surf the web” was an extremely effective marketing message and drove the site to be one of the top-twenty trafficked web properties.  The service required users to view advertisements targeted to them.  AllAdvantage split the advertising revenue with its users.  As you probably guessed, AllAdvantage is no longer around today.  Unfortunately, it’s hard to know exactly why AllAdvantage failed.  It was definitely caught in the internet bubble having raised over $200 million in venture capital, and was hit hard by the sharp downturn in 2002.

Online Video

Other companies employing the less-than-free strategy include video sites Revver.com, Shareaflick.com, Metacafe.com, and Blip.tv.  With these video sharing sites, not only can users post video’s for others to see for free, they can profit if the video generates a significant audience.  Similar to AlAdvantage, these sites offer revenue sharing such that the advertising revenue that their site generates is shared with the producers of the content.

Does Less than Free Work?

Revenue sharing seems to have potential as a strategy for disrupting incumbents in the consumer web space, but it’s hard to point to any case where it has actually worked.  One possible explanation is discussed in the book “Predictably Irrational” by Dan Ariely.  Ariely’s research shows that there is often a huge, irrational increase in demand created by decreasing the price of something from $0.01 to $0.00.  This may explain why so many web-services have generated huge audiences by giving their services away.  However, Ariely also shows that paying people to do something can decrease the effort they put into doing it, especially if the amount they are being paid is relatively small.  Why?  When paid to do something, people tend to work as hard as they feel the payment justifies.  When we do something because we are intrinsically motivated, we tend to put in our best effort.

Although it’s hard to point to an example of less than free being a successful disruptive strategy, I hesitate to totally count it out.  Perhaps in the right context it could give one company the ability to surpass its rivals, particularly in situations where the performance difference between competing products is small.  For example, if Yahoo held a promotion where it offered a daily $100,000 raffle to anyone using its search engine that day, it could probably win many of Google’s users.

Low-Time Disruption

Low Time Disruptive Innovation

Another strategy to disrupt a free-web service is to change how you think about the price of a service.  Rather than trying to compete on the dollar cost (which is already zero), focus on the minute cost.  For many consumers, time is of more value than money anyways.

As Figure 3 illustrates, a disruptive strategy could be to undercut competitors with a service that takes less of a consumer’s time to get something done.  The disruptive service can have less flexibility or less features than the incumbents, but must still provide some value that the customer desires.

Facebook Disrupting MySpace

One example of a free web-service disrupting another with a “low-time” approach is Facebook’s rise over MySpace.  As shown in Figure 4, while MySpace enjoyed a comfortable lead as the first social network to reach 20 million users, Facebook has continued to grow its user base and by some metrics surpass MySpace .  Many differences between the two sites have been attributed to this disruption, but the “low-time” disruption model offers a compelling explanation.
In order to keep track of friends in MySpace, a user must browse their personal pages to see what they are up to.  To keep track of who is dating who, where friends are travelling, or what someone did that day, users spend hours clicking through all their friends pages, reading the messages others have written, viewing pictures and generally staying in touch.  The effort provides huge amounts of information, but can be an all day affair.  In the disruptive framework, this is like paying a high price for a product that (may) exceed a customer’s needs.  Consumers that find the process “too expensive”, end up not spending the necessary time to keep track of the majority of their friends, and thus receive little value from MySpace.

Facebook entered the social networking space later than MySpace and did not begin to see large customer base growth until 2007, when MySpace already claimed over 20 million users.  Regardless, Facebook was able to surge past MySpace.  To explain this disruption, consider Facebook’s “News Feed”.  The News Feed is a core feature of Facebook, central to a user’s home page, and one of the first things seen when logging in.  The News Feed displays, in reverse chronological order, recent events that have occurred in one’s Facebook network.  With the News Feed, it’s no longer necessary to browse through all of one’s friend’s personal pages to see if anything new has happened in their Facebook life.  Recent events such as posting new pictures, making new friends, updating a status, or commenting on others pages are consolidated and displayed right on the user’s home page.

Social Network Market Diffusion

In many ways, Facebook is less of a product than MySpace.  It lacks the ability for a user to customize profile pages with music, videos, and stylized html that MySpace provides.  Browsing around all your friends MySpace profiles definitely provides more information about one’s friend’s lives and personalities than simply glancing over one’s News Feed in Facebook.  However, not everyone desires all the information that MySpace provides.  Just as the extremely large capacities of rotational hard discs were not required by all computer users, the large amount of information provided by MySpace is not required by all social network users.  Facebook provides us slightly less information than MySpace, but with a huge time discount, demonstrating the low-time disruption strategy.

Twitter Disrupting Blogs

According to Wikipedia, a blog is “a website, usually maintained by an individual, with regular entries of commentary, descriptions of events, or other material such as graphics or video”.  Popular blogging platforms include Wordpress, Blogger and Movable Type.  Blog platforms are much easier to setup than building a website from scratch, and have in many ways disrupted web development software such as Dreamweaver and Front Page.  Today, there are reportedly over 100 million blogs.

Even Blogs may be disrupted with a low-time strategy.  The blogging platforms described above give the blogger a significant amount of flexibility.  Users are able to modify the source code of their site, manage comments, publish additional static web pages and embed an endless offering of widgets in the margins to provide added functionality.  Blog posts are typically 200-600 words, with some being much more.  Writing a good blog usually involves thinking carefully about the post, referencing sources with links, spell checking, fact checking, and even adding some insightful commentary to the facts being reported.  This can be a very time consuming process.

Twitter, described as a “micro-blogging” service, uses a low-time approach to disrupt other blogging platforms.  All Twitter posts are limited to 140 characters long.  Given this strict limitation, it’s impossible to write well structured commentary the way one might with a blogging platform.  However, this seems to be plenty of room for “Twitterers” to get their points across.  With traditional blogging platforms, some customers were “over-served” with too much capability, and unable to afford the time necessary to manage a full blog.  Twitter provides a less capable product that takes less time to use and thus also follows the low-time disruption strategy.  With only 159,000 monthly unique visitors, Twitter still has significantly fewer users than blogging platforms such as Wordpress, which reports over 151 million monthly unique visitors.  Twitter’s user base is growing quickly though.

Posted in Internet, Social Networks, Startups | 1 Comment »

Value Capture, Complementary Assets and Uniqueness in Social Networks

March 21st, 2008 seph250

There’s no doubt that with the fast rise of social networks, immense value has been created. Myspace now ranks as the fifth most popular website globally and Facebook ranks seventh. It’s unclear, however, if these services will be able to capture this value in the future.

Friendster

As one of the first social networks to enjoy success, Friendster is an awesome example of failing to capture value. Today, Friendster sees considerable traffic, ranked the 17th most popular website globally. However, in December 2007, Friendster saw revenues of $700,000, with a loss of $400,000. Why, with so many users is Friendster still not being profitable? The answer is in Friendster’s user base. Today, 70% of Friendster’s traffic comes from the Philippines, Malaysia, or Indonesia. Only 4% of traffic comes from the United States. With such little traffic coming from wealthy countries, online advertising (key to social networks revenue) is ineffective and Friendster is unable to capture value.

Myspace

Friendster’s founder Jonathan Abrahms is credited with developing the idea of an online social network, however, without any real intellectual property protection, or secrecy, the idea was not unique for very long. It would seem that Myspace took the same idea and built the first successful social networking business, thus creating and capturing value. MySpace’s web pages are crammed with advertisements and MySpace is profitable. According to News Corp, MySpace, turned a profit of $10 million on revenue of $550 million for the fiscal year ending June 30, 2007. MySpace revenue is projected to surpass $800 million in fiscal 2008.

Facebook

Facebook has also created value with its social network, however so far it has captured very little. In 2007, reported revenues were $150 million, with an EBITDA of $50 million. Projected revenues for 2008 are $300 to $350 million. Facebook has only recently begun to insert advertisements in its site, and it has been very judicious in how it goes about doing this. As opposed to MySpace, Facebook shows fewer advertisements per page and they are generally more subtle.

Complementary Assets in Social Networks

Complementary assets are critical to the success of social networks, which inherently rely on network externalities to be successful. The following complementary assets play a large role in the competition between social networks and will continue to do so in the future.

User Base

Social networks are all about the people that use them. A person is most likely to join a social network that his/her friends are already in and a positive feedback loop occurs. There are also switching costs (although small) of moving from one social network to another so a new customer for one social network often means one less potential customer for a competitor.

Servers

In 2008, Facebook plans to invest $200 million in new servers. Increasing its number of servers and having servers located across the country is an expensive upfront cost, but it will allow Facebook to deliver faster loading pages to a broader audience. If other social networking sites do not take action similar to this they may be at a disadvantage. As video and interactive content on social networking sites increases, these servers will allow Facebook’s pages to load faster and deliver an improved user experience.

Third Party Applications

Facebook was the first to announce the Facebook Developers Platform, allowing third party developers to build and launch their own applications within Facebook. Since this announcement, Myspace and Friendster have made similar announcements. By allowing third party applications, social networks hope to create another complementary asset of third party developers and rich applications for users to benefit from.

Interestingly creation of this asset is another example of a social networking site passing up an opportunity for value capture. Instead of developing these applications internally and capturing all the advertising dollars, Facebook allows third parties to profit. In the first six months of launching the platform, third parties are estimated to have earned $20 million in revenue.

Uniqueness in Social Networks

There is no defensible IP in social networks (that I know of anyways). Theoretically any social network could copy another. Many startups today have taken what seem to be the best features of Facebook, MySpace and others and launched (arguably) better social networks than either of the leaders. However, without the complementary assets these startups have gotten little traction.

But, uniqueness still plays a role in Social Networks. With such a large install base (so many users) it is not always possible for social networks to copy each other. For example, it is a fundamental feature of MySpace that users can customize the HTML of their page. So, suppose MySpace decides that they prefer the clean, organized look of Facebook. To make that change they would have to make major changes to the system and run the risk of angering and losing many of their current users. Thus, they are forced to stick to more evolutionary changes to the system as demonstrated when MySpace launched a news feed very similar to that on Facebook.

Uniqueness has and will continue to play a major role in completion between existing Social Networks. The subtle differences between Facebook and MySpace may cause one or the other to ultimately prevail.

Posted in Internet | 4 Comments »

Market Diffusion of Online Social Networks

March 5th, 2008 seph250

Online Social Networks have seen huge success in the last five years, going from less than one million users in 2003, to an aggregate of over 100 million users today. This market diffusion has been fueled by broadband penetration, mobile access, international expansion, product improvements and the obvious network externalities. This recent diffusion has changed the entire industry, shifting power from traditional internet leaders (Microsoft, Google) to the social networking sites (Facebook, MySpace). In this chart here, I am simplifying data from Alexa and inferring the number of uniques users from comscore reports, and quantcast data. I’ll consider the number of unique visits as the number of “customers” these sites has.

socialnetworkdiffusion.jpg

Broadband Penetration Increasing

A major external factor limiting market diffusion of any product is distribution footprint. Since social networks are an Internet product, let’s consider internet access as a measure of a sites distribution potential. Broadband penetration in the US has increased from 33% in 2003 to 85% in 2008 , allowing increased distribution of online social networks. This growth has been a major fuel for social network diffusion.

Improvements in Mobile Access

Ease of access to online social networks via mobile devices has also improved dramatically in the last few years. In February 2006, Helio launched a phone featuring MySpace Mobile that allows users to perform the most popular MySpace functions. Another milestone was the launch of Apple’s iPhone in June 2006. The iPhone’s improved browser interface makes interacting with a social network much easier.

However, not everyone wants to access their social network on the go. According to Jupiter Research, only 2% of cell phone users use their mobile device to access an online social network while 12% of social network users are interested in accessing the sites on their phones. However, for some people without broadband, mobile access is their only way to participate. To capitalize on this, in September 2007, MySpace announced plans to launch a free ad-supported mobile phone. If they pull this off, it would open MySpace to users without broadband or mobile access today and grow their user base considerably.

Product Improvements

Another major factor influencing diffusion of social networks is the continuous improvement of the social networks themselves. Features such as content creation; flexible customization of pages and the ability to share media are adding greater utility for consumers and ultimately drawing more people to social networks. For example, at the time of launch of all three major social networks, there was no method for a user to upload and share videos. However, MySpace launched a video sharing functionality in January 2006 , and Facebook launched the feature in May 2007.

International Expansion

Major social networks are taking advantage of the scale of the Internet to diffuse their products and to create value beyond the US. In 2006, after being acquired by Fox Interactive,

MySpace expanded into 15 foreign countries, developing local versions of the site and marketing the site internationally. Currently, MySpace has developed local versions of its site for 24 different countries.

In fact, despite Friendster’s decreasing popularity in the US, it grew substantially in Asia in 2007, which now account for 89% of its unique monthly visitors.

Industry Evolution

Shift in Market Power

With the increased diffusion of social networks, the dynamics of the industry have changed dramatically. Just five years ago many people regarded social networks as a fad and waste of time. However, today with Myspace and Facebook producing so much traffic, they are regarded as extremely powerful companies. An indicator of this is when in October 2007 Microsoft paid $240 million for just a 1.6% stake in Facebook.

Transformation from IT Companies to Media Portal

When social networks first emerged, their primary focus was on communications and networking. However, in the last few years the sites have evolved to be more like media portals. For MySpace, this was driven partly by its acquisition by News Corp, a major media company, in 2005.

The result is that these social networks are now competing, not just with each other, but with more traditional portals like Yahoo, AOL and MSN, as well as media sites such as YouTube and Flickr.

Transformation from a Walled Garden to a Platform

Another important aspect of the industry evolution is that Social Networks are opening up their APIs and allowing other companies to develop and market applications that run on them. Facebook announced the Facebook platform in May 2007 and as a result over 5000 companies and independent developers are currently developing applications that run on Facebook.

This changes the industry, because now these developers no longer look at Facebook as competition but as a powerful ecosystem within which they can profit.

In November 2007, MySpace committed to the Open Social standard developed by Google allowing third parties to develop applications that run on MySpace and other social networks such as Bebo, Ning, Plaxo and Six Apart. Similar to Facebook, Myspace has also announced a developer platform.

Overall, the online social network industry is rapidly growing and dynamic. The factors driving market diffusion, such as broadband penetration, improving mobile access and product improvements are themselves continuing to evolve. The market structure and industry evolution indicate that social networks are poised for continued growth as the late majority begin to create their profiles.

Posted in Internet | 1 Comment »

Top Six Core Competencies for any Internet Startup

March 1st, 2008 seph250

One of the coolest things about going to business school at MIT is the General Catalyst Breakfast Series. Once a month, 10 MIT students and 10 HBS students wake up painfully early to go to General Catalyst’s office in Harvard Square to listen to a speaker give their take on entrepreneurship. Last Friday, we saw Paul English, the CTO of Kayak.com. Several people asked for my notes, so I thought I would post some of the highlights. Obviously, these are all Paul’s words, or at least how I interpreted them.

About Kayak

Kayak was originally conceived as the “google of travel”. Today, it competes with Travelocity and Expedia. Kayak does about 34 million searches/month while Expedia and Travelocity do 150 million each. Kayak is unique in that it is not an Online Travel Agency (OTA). Kayak does not book the flights for you, it just forwards you onto the site with the best price. It really is a vertical search for travel.

Startup Core Competencies

Paul described what he feels are Kayak’s six core competencies – but it’s interesting to see that these core competencies can be applied to any internet startup. I’ll list them and then elaborate on the first two.

  1. Hiring/Firing
  2. Interface Design
  3. Customer Interaction
  4. Business Development
  5. Industry Knowledge
  6. Fund Raising

Hiring/Firing

Kayak produces $150 million in revenue with just 50 employees. $3 million revenue per employee is the result of selective, aggressive hiring. Kayak now has 15 developers and only one UI designer. This is a site that manages 34 million searches per month. Paul says he looks for four things when hiring someone:

  1. Bandwidth – only hire the best of the best. Super programmers can produce software an order of magnitude higher than average.
  2. Attitude – Look for people that are extremely confident. These people are able to give credit to others and lift the whole team up. Also, look for people that are “Energy Amplifiers”. These are people that draw in other people. Kayak moves at “Kayak Speed” (extremely fast) without working nights and weekends. An example of this is that Kayak submits a new version of their product on a weekly basis.
  3. Experience – Look for people that have a history of being extremely successful. As an example, Kayak employs an Olympic gold medalist! One thing I found interesting about this is that Paul does not look for industry experience. He has no interest in “domain knowledge” – as though this sort of experience will corrupt his thinking…
  4. Lack of Dysfunctional Behavior – He didn’t elaborate on this much, but I think it’s pretty straightforward: No major assholes. 

Interface Design

Maintain an emphasis on site speed. If a new feature slows the site down, get rid of it. Paul talked about what he calls the 5 “B’s” of UI design. These are the things a user looks for when they get to your site. Providing them in a convenient way will provide for a more intuitive experience.

  1. Bikini – This is the first thing a user looks for – either something pretty, or the data that I am looking for.
  2. Button – If I don’t see the Bikini, I am looking for a button I can click that will take me to the Bikini.
  3. Blue Link – If there’s no button or bikini, I am looking for some blue text that I can click on to take me where I want to go.
  4. Box – If I don’t see any of the above, I am looking for a box that I can type into. Enter in my email address, or name, or type my search query.
  5. Bullshit – Finally, everything else on the page is just bullshit and it is the responsibility of the designer to minimize the bullshit. Examples of bullshit is text that says “Click here to get the best deals on flights to Hawaii”. A website that needs text to explain how to use it should be redesigned.

Posted in Internet, Startups | No Comments »

Social Networks and Gossip Distribution

February 12th, 2008 seph250

The term “social network” is pretty abstract with today’s social networks providing diverse ranges of functionality including entertainment, personal messaging, media sharing, self-expression, networking and gossip distribution. By “gossip”, I don’t mean negative or slanderous talk about others. More generally, I am referring to any casual information about other people in our social circles lives. Examples include someone going on a trip, getting a new job, breaking up with their girlfriend, or even getting a new hobby. Gossip distribution, which most people love to do, is a unique type of communication. While most forms of communication direct us to do something, or deliver information critical to our lives, the purpose of gossip distribution is to create a general awareness of what is going on in our community and understand each other better. I’d like to evaluate the historical methods of gossip distribution and try to understand how social networks have become a tool we rely on for this function

Industry S-Curves

Historically, the technologies and methods used to deliver gossip include:

  1. Physical Meetings
  2. Telephone
  3. Email
  4. Social Networks

socialnetworkcurve.jpg

In Figure 1, I outline the performance of the different technologies over time. The x-axis represents time and the y-axis is a performance metric titled “Ease of Information Distribution”. We mathematically define this metric as (number of relevant friends reached)*(quality of information) / (time spent distributing the information). Clearly I am not considering the scale of time or magnitude of performance improvements, but I think the general concept of these technologies disrupting another as our primary tool is reasonable. I am also ignoring entire categories of technology used for gossip distribution for the purpose of simplification.

Physical Meeting Points

Going back 1000’s of years, communal points provided the first method of gossip distribution. Ease of information distribution was low because delivery required physically meeting and the maximum number of people that could be reached at once was limited. Potential performance increases were possible with new technologies such as the megaphone and amphitheater however these were not well suited for the distribution of certain types of personal gossip information and could not always be utilized.

Telephone

With the advent of the telephone, the ease of distributing gossip dramatically increased. Physical travel was no longer necessary. As three-way calling, mobile phones and speaker phone technology became available, the ease of gossip distribution further improved.

Email

With email came a new method of distributing gossip. Users were now able to create mailing lists and send messages to a very large number of targeted friends simultaneously. As network externalities improved and more and more people became comfortable with this communication medium, performance further improved. Improved email software and email available on mobile phones also increased the performance. Users of email found that composing a message to deliver gossip information was more work than just talking on the phone, but being able to send it out to several targeted friends simultaneously made it ultimately less work.

Social Networking

Finally, social network sites emerged proving to be the ultimate (thus far) tool for gossip distribution. Early versions of social networks were bulletin boards, forums, AOL communities, and Geocities websites. However, it was not until later versions of social networks such as Friendster and Myspace that social networks began to outperform email as a gossip delivery tool. The design of these social networks allowed a user to browse their friend’s profiles and learn gossip (updates) about the person by viewing their photos, reading their “status” and reading their blog postings. Another key feature was the “comments” section in which users posted messages to each other that were visible to all other friends.

Hence, the “Social Graph” emerged: relationships began to be represented online with greater clarity.  Distribution of rich gossip information became automatic.

The person providing the information was simply required to post the information and anyone interested could (with a little digging) view it. Finally, this technology was further disrupted when Facebook developed the feature called “news feed”. With the news feed, users are no longer required to actively browse their friend’s pages to get updates – instead they are automatically filtered and delivered right to the person’s main page.

Past and Future Disruptions

Thus far, many disruptions have occurred in the gossip distribution industry. At a high level, we see in-person meetings disrupted by telephone disrupted by email disrupted by social networks. At a lower level we see that there is disruption occurring within categories such as the Friendster – Myspace – Facebook evolution. Even with Facebook’s news feed, we see significant room for improvement and thus opportunity for disruption. Social networks will likely evolve to distribute more and more gossip information with minimal incremental effort by the distributing person through increased integration with other online information sources. As more information becomes available, technology will need to improve such that it is more selectively distributed to interested persons so as to avoid a flood of information. Perhaps disruptive improvements can be made in the method of delivering this information. Improved distribution through mobile devices and other electronic media will allow users to share and access gossip information easily and on demand. Many companies are currently well posed to disrupt Facebook with some of these improvements.

Natural Technological Limits

One potential natural technological limit to these technologies may be rooted in our humanness. In many cases we continue to prefer to distribute gossip through in person discussions. Social networks and other online systems provide a breadth of information with very little effort, but have so far failed to yield the same depth of gossip distribution that is possible with in-person interaction. Perhaps technology, even with video or virtual reality will be unable to replicate face to face dialogue… or perhaps it will.

Posted in Internet | 1 Comment »

Advertising Causes Social Network Peak?

February 8th, 2008 seph250

Business Week has an interesting article Generation MySpace Is Getting Fed Up suggesting that social networks are being hurt by attempts at monetization. It has some interesting data points to analyze:

The average amount of time each user spends on social networking sites has fallen by 14% over the last four months, according to market researcher ComScore.

I’m skeptical why this statistic is based around 4 months… Was there an odd spike last October? Are these numbers really going down or just fluctuating? Is 14% really that significant? Business Week builds a case that this decline is caused by the increase in advertising, but there are many more things going on that could cause this. When a user first signs up for a social network, there is a novelty aspect to it and the user spends massive amounts of time exploring. However, once you’ve found all your “friends” and stylized your page to your liking, there’s simply less to do on the site. Sorry Business Week, I disagree. This decrease in time online is not due to the advertising - We just became bored with the concept. Another datapoint:

MySpace, the largest social network, has slipped from a peak of 72 million users in October to 68.9 million in December, ComScore says.

First of all, let’s be clear what these numbers are referring to. ComScore bases these numbers on the number of unique visitors to a site over the course of the month. ComScore does not know how many registered users Myspace has. This is not about users canceling their accounts, this is about Myspace users not going to Myspace in December. One argument is that this decline is caused by Myspace users switching to Facebook, or some other social network platform that they like better (I tend to believe this). Another explanation (that Business Week argues) is that users leave the social network site because they get too annoyed with the advertising. But wait a minute, we’re talking about December! This is holiday season when everyone goes on vacation. If you’re not at your desk at work looking for ways to goof off, then you’re not going to MySpace… And like I mentioned before, we’re getting a little bored with what Myspace has to offer and don’t obsess over our accounts the way we used to.

Bottom line, the way I see it, Social Networks are not in bad shape, they’re just beginning to mature. Even Business Week acknowledges

The total number of people on such sites is still increasing at an 11.5% rate, but that’s down sharply from past growth rates.

And it’s not the advertising driving us away, it’s boredom. Social Network sites can’t expect to build an asset and milk it for perpetuity. Constant innovation and improved services will be necessary to keep users attention.

Posted in Internet | 1 Comment »

Microsoft Yahoo Merger Good for Startups!

February 6th, 2008 seph250

Marc Andreesen, successful entrepreneur (Mosaic, Netscape, Loudcloud) and investor (Digg, Plazes, Ning) provides a nice analysis of the potential effects of a Microsoft-Yahoo acquisition to the likelihood of smaller startups being acquired. His conclusion is that there will be no effect :

The Microsoft/Yahoo deal, if it happens, means very little for the entrepreneurial climate in Silicon Valley, or the opportunities available to you and your startup.

His reasoning (to summarize) is the following:

  1. Yahoo and Microsoft weren’t buying that many companies anyway.
  2. There is a broad spectrum of companies making acquisitions.
  3. Traditional media companies will make more acquisitions in the next two years.
  4. Entirely new companies can emerge quickly and make acquisitions.
  5. Building a startup with the goal of being acquired is a stupid idea anyways.
  6. The practicalities and logistics of a merger would prevent Yahoo and Microsoft from aggressively developing new products and services, improving opportunities for startups.

All in all, he generally paints an optimistic picture for the consumer internet space.

I’m pretty confident guessing that the level of concern and even panic among many traditional companies — particularly media companies — is only going to escalate from here, as traditional non-Internet businesses in various sectors deteriorate and consumers continue moving en masse to the Internet.

And from there, it’s not hard to guess that Internet M&A is likely to heat up considerably over the next several years, compared to the last several years, across a very interesting and surprisingly diverse cross-section of buyers.

While many people are talking about a web 2.0 bubble, I tend to feel that the general trends look very good for consumer internet. By looking simply at the scale of the rise in online spending, and broadband penetration, I am confident that there will continue to be many opportunities for startups to be successful. The internet is here to stay and there are still miles of room for innovation and improvement.

By the way, here’s another interesting post discussing the Microsoft Yahoo Merger suggesting an alternative to a full fledged acquisition…

Posted in Internet, Startups | No Comments »

Strategy: Using Illegal to get to Legal?

February 5th, 2008 seph250

Ashwin Navin, the president and founder of Bit Torrent Inc. gives a podcast describing how Bit Torrent Inc was started. Bit Torrent was originally an open source peer to peer file sharing protocol developed by Bram Cohen to deliver large amounts of data over the internet quickly. Initial users and evangelists of the protocol were developers distributing Linux. It was a perfect fit: Linux and Bit Torrent were both open source projects so the philosophies resonated, and Bit Torrent allowed Linux to be delivered quickly and cheaply.

However, Bit Torrent distribution really went through the roof was when it started being used for illegal sharing of movies and music. Millions of users across the web (including myself) wanted to get free mp3s. By installing the Bit Torrent client, we were able to download them quickly. It’s an interesting situation to consider:

You are developing a product with the best intentions, but your product is being used to break the law…

Of course, this was a law that most of us felt no moral obligation to… Nobody seemed to be getting hurt and it weas hard to feel bad for the pop-star millionaires complaining that they sold less records.

At some point, Ashwin left Yahoo to start Bit Torrent Inc. The basic idea was to leverage the broad distribution of the Bit Torrent client into a legitimate business. In his podcast, he describes his pitch to build relationships with large media companies:

Mr. Studio Chief… In the next hour that we’re going to be talking, about 70,000 of your movies are going to be downloaded illegally with my software… We got to do something about that…

Pretty hilarious! And for a large part it worked. Bit Torrent may be the closest competitor to iTunes in legal downloads. This is definitely a strategy not taught in Business School, but worth considering as I’d expect it to be repeated in many more new media companies. Consider the discussions online video and music sites such as YouTube, Pandora and Musicovery are having with studios and record labels. - we’re helping users get your content for free - let’s work together to monetize this…