"Integrating with Frengo was a breeze and we really appreciate their experience and expertise in mobile. With Frengo we can extend our social experiences to people on mobile phones around the world," said Jia Shen, CTO of RockYou, a launch partner with Frengo on the toolkit. RockYou's "Horoscopes" application is available to mobile users via the Frengo toolkit. Other customers of Frengo include Slide, I Can Has Cheezburger?, Serious Business, and Frozen Bear.

Lance Takuda of RockYou recently confirmed to us that there are slight differences in the Open Social deployments on MySpace and Hi5, and the Facebook platform deployments on Faceook and Bebo. He told us there's about a "20% overhead in supporting" the different deployments of each platform (though going platform-to-platform basically requires a rewrite). It seems likely that because of these differences, the Frengo toolkit includes slightly different bits of code for working with each social network.

It is, of course, not out of the realm of possibility that social networks could build mobile functionality directly into their platforms. Both Facebook and MySpace have been pushing their mobile versions hard recently -- MySpace just partnered with Sprint and Verizon and RIM just announced a million Facebook users on Blackberry -- and it is plausible that they could push developers to the mobile space themselves. Facebook especially has some serious mobile chops with Joe Hewitt on staff, whose iUI is already one of the most popular iPhone frameworks.

Over the past month, we've continued to think a lot about the process of picking a platform on top of which to build an application. We've had a few interesting interviews on ReadWriteTalk that have touched on these issues. Specifically, we interviewed David Glazer, an Engineering Director at Google and one of the champions inside Google of the OpenSocial project. We also interviewed Lance Tokuda, the CEO and Co-Founder of RockYou, a social software company that builds apps on top of OpenSocial and the Facebook platform.

Though the interviews with Lance and David focused on the issues related to platforms for social applications, it struck us that the same dynamics now exist across a number of platforms. For example, in the mobile space there are a number of platforms on which you can currently release a mobile app. We actually covered some of this in our interview with Adam Taggart from Yahoo!'s Mobile Platform. Obviously, with Apple opening up the iPhone platform, the alternatives are growing. Another industry with similar dynamics is the video game market. In fact, Lance actually referenced Electronic Arts as a model for the type of company he is building.

In this post, we'll highlight five common dynamics that software companies need to take into consideration when deciding which platforms to focus on.

Audience Issues

A platform will only be of use to you if the type of user who will find your application valuable is actually using it. Therefore, the first three issues to be considered when picking a platform center around the dynamics of reaching those users and making sure they are the right users for you. Specifically, these are:

• Issue 1: Reach of the Platform
• Issue 2: Growth of Audience
• Issue 3: Value of that Audience

Issue 1: Reach of Platform

Reach is simply another term for the size of an audience. When looking at reach, it's important to consider how many users specifically have the problem you're trying to solve. If a platform doesn't reach your target audience (and you don't anticipate that changing), it is a non-starter to even consider building on top of the platform. However, platforms that efficiently allow you to reach your target audience are very desirable.

For example, at all the tech conferences I've attended over the last year it has been striking that a high percentage of the audience are iPhone users. In one session at SXSW this year, the panelist asked the audience to raise their hand if they owned an iPhone and probably 75% of the crowd raised their hand. If you had an idea to build a mobile application that was targeted at innovators and early adopters in the technology industry, the iPhone market might actually be the audience with the greatest reach for your app. I would argue this is true even if other platforms have more total users right now.

On the other hand, if an application had a broader target market then other platforms might make more sense. That is clearly the type of application that Yahoo!'s mobile platform is targeting. Adam Taggart shared the following statistics in our interview:

We are now in the process of developing a mobile ecosystem that is intended to serve eventually billions of mobile consumers. And I had to use the word "billions" there very specifically because it differentiates our strategy versus other people in the industry. Most of the people who are taking a very highly optimized approach around the specific device type or a specific operating system, which is great ... So if we'll take the iPhone for example, you know, a phenomenal, fantastic device and very, very popular. They will probably sell in the middle of this year their two millionth iPhone ... So while you're promoting a very good experience, you're providing really just a very thin sliver of the total market of consumers out there with phones.

Issue 2: Growth of Audience

One of hockey great Wayne Gretsky's most famous quotes is, "I skate to where the puck is going to be, not where it has been." In many ways, the art of choosing a platform is very much the same. Companies need to be more focused on the platform that will be most valuable in the future, instead of picking what is most popular at the time of development.

In social applications right now, an interesting aspect of this will be web based email clients. In our interview, Lance Takuda expected that major services would end up leveraging OpenSocial versus using the Facebook Platform or building another platform. Listening to Lance talk about the 250 million users that web based email clients represent, it was clear that RockYou has aspirations to dominate this space. I would guess that rival web app maker Slide is targeting it as well. However, I also anticipate companies focused on improving the email experience, such as Xobni to compete aggressively once the web based email clients go live.

Of course, while web based email platforms serve as a good example, the overarching point is that if you are designing and developing software on a platform, make sure you're building for where the users are going to be, not where they have been, to give your application the best chance of long term success.

If you're really good at anticipating where users are going, you can end up with what is often called the "first mover advantage." RockYou is a great example of this. They were one of the early companies to recognize the significance of the Facebook platform and released numerous social applications as soon as it opened up. They were thus able to leverage that quick response to become 'the fastest growing company in the world in terms of user acquisition ... basically hav[ing] grown to 60 million users in about 26 months.'

Issue 3: Value of that Audience

Interestingly, 'value' can be looked at in a number of different ways. For example, RockYou focuses on an audience's value to advertisers - specifically, in their case, looking at the geography of where users live. To an advertiser, a teenager in the United States is high value demographic, so that is one of the reasons they are specifically building apps for US teen heavy social networks.

However, as we recently pointed out free to end users is sometimes a dangerous approach. You might want to think about the service fees different audiences on different platforms might be willing to pay directly. If you do, it's important to consider how efficient the payment mechanism is. For example, Facebook is building a payment system as an integrated part of their platform.

Technology Issues

In addition to the issues around the audience, you can't overlook the pragmatic issues around technology. Specifically, there are two technology factors you need to take into account:

• Issue 4: Capability to Support Desired User Experience
• Issue 5: Efficiency of Developing on Platform

Issue 4: Capability to Support Desired User Experience

When you build on top of a platform, you often inherit a set of limitations that are either business decisions by the creators or simply technical limitations based on the way it was developed. While we are very excited about the applications coming out with the upcoming iPhone SDK, it is interesting to look at some of the limitations that SDK imposes upon developers.

For example, according to Wired, developers are not able to create programs that continuously run as a background thread. This clearly limits the functionality a group creating an instant messenger tool might be able to create. In this case, I believe it's unclear whether that is simply a technical limitation or a business restriction. It could be a business limitation if Apple were planning on releasing their own chat application, as some have speculated.

On the other hand, the iPhone also has certain capabilities that previous mobile platforms did not, which is why we believe that Apple will dominate next generation computing.

Issue 5: Efficiency of Developing on Platform

Another important issue to take into consideration is how efficiently your team is able to develop its application. Specifically, this takes into account things like how well documented the platform is and what tools have been created to make development more efficient. David Glazer pointed out there is an interesting effect of having a compelling audience (the three issues above) on these tools and documentation being created:

There's a virtuous cycle that goes on where as a particular environment has high reach, it attracts a lot of developers. When it attracts a lot of developers, there's a lot of value in creating the tools to make it more efficient to develop for that platform.

It is also important to take into consideration any situation where there may be slightly different deployments of a general platform. For example, the OpenSocial platform is slightly different on MySpace than Hi5 and the Facebook platform being leveraged by Bebo is also slightly different. Lance Takuda commented:

For example, between MySpace and Hi5, there's only about 20% overhead in supporting both. Whereas from Facebook to MySpace, there's almost a rewrite involved.

He also confirmed that it is a similar 20% overhead in supporting an application across the Facebook and Bebo platforms.

Conclusion

When you look at RockYou's rapid user growth over the past 26 months, it's obvious that people can build successful businesses on top of different platforms. However, it is critical that developers evaluate the right platform for their application. We've laid out a list of 5 critical dynamics to consider. What factors did we overlook or which dynamics should we have highlighted? Please let us know in the comments below!

[Note: This post drew heavily on three recent interviews on our podcast, ReadWriteTalk. If you'd like to listen in on future interviews, please consider subscribing in iTunes or via your favorite RSS client.

Image credits: Audience Photo & Technology Perspective & The Great One

GSP kicked off with the Charlene Li keynote. Charlene talked about the social graphs provided by the Facebook Platform and OpenSocial as "the air of tomorrow's web." What she meant was that the social graph is going to be an essential piece of tomorrow's web, and the sites which don't embrace it will most probably become extinct.

The second keynote came from Amit Kapur of MySpace. Amit gave important insight about the monetization of their network, but admitted that "social networking is hard to monetize," which is something we have also heard from Google. Amit said that regular contextual advertising doesn't work well for social profiles and that's the reason why MySpace is focusing on their own monetization technologies, which they call, HyperTargeting. HyperTargeting is already in use by big clients such as EA, Microsoft, Taco Bell, Proctor & Gamble, and FOX. But he added that they are also planning to release a self-serve solution for long-tail advertisers. Presumably, it will be a social doppelghanger of Google AdSense.

The third keynote was given by Benjamin Ling of Facebook. He made some important announcements, including the new E-Commerce APIs that they are planning to release soon, as well as the new user-driven localization solutions that they will provide for app makers. Facebook Platform investors at the event were unanimous in their belief that the E-Commerce APIs are big news, because it means new business models and revenue sources for app producers. On the other hand, their localization solutions seem to be just another step in their recently debuted internationalization efforts. Indeed it would be inconsistent to offer a localized platform which barely has any localized apps on it. And the good thing is, since this is user-driven, the costs for app makers to get international is low and also linearly dependent on their merits.

Day One Announcements

Following the keynotes, there were a number of interesting announcements made, including:

• NetVibes introduced Ginger. Ginger is a social version of NetVibes that allows you to pull in friends from your existing social networks, share news with them and follow their reading activities. In other words, as NetVibes' Chris Damsen noted, Ginger makes Facebook, your private place on the web, a more public place.
• MyBlogLog introduced Bluetooth capabilities and a new FriendFeed-like feature. The Bluetooth feature allows you to see other MyBlogLog users who are close by. When we tested it with Ian Kennedy of MyBlogLog, at least 10 MyBlogLog users who were attending the conference showed up on our list. In some sense, this new feature brings MyBlogLog's distributed social networking into the real world.
• Bebo, Myspace, hi5, and Friendster are launching or have already launched their own OpenSocial-based platforms.
• Chris Messina of Citizen Agency introduced his distributed social network project, DiSo. DiSo's architecture depends on existing open standards such as OpenID, XRDS-Simple, and microformats. It sounds very geekish for now and is too-focused on technical aspects, instead of the social realities that actually make a network work. But it has the potential of becoming the "Linux" of social networks.

Day Two

Compared to day one, day two began with more technical topics. MySpace's Jim Benedetto gave some information on their new platform, which extends existing OpenSocial v0.6 standards with MySpace specific features like bulletin boards. The interesting point about the MySpace Platform is that it is going to launch very restricted. All app submissions will have to go through a safety review process by humans. Hence there is going to be an unavoidable lag between releasing an app and its availability on the network. But all these limitations will be slowly removed with a measured approach, according to Jim. He said that this is to prevent spam and protect the long-term value of the platform.

As most of the app developers in the conference noted, including representatives from RockYou! and Slide, restrictions by platform providers will definitely cut down their fast viral growth opportunities, but the applications that create a real value for the users will still be able grow virally, perhaps at a slower pace but stickier.

One of the shortest but definitely most informational sessions of the event came from Roger Magoulas of O'Reilly Media. He showcased some very interesting numbers and statistics that O'Reilly has collected from the Facebook appsphere. In summary, he showed that:

• installation numbers are no longer as big as they used to be
• active usage rates have dropped significantly too
• most of the apps are released under the "just for fun" category
• there is more adoption and engagement in "games"
• there is a tendency towards winner-takes-all; 1% of apps have 75% of app users, 20% have 99%

Later, in a session that brought together top Facebook game developers, including Mark Pincus of Zynga, there was agreement that social games will replace casual gaming, just like Facebook messages are replacing emails. Also Mark noted that live games don't work so well on Facebook, because once people leave the game, they don't come back; so in order to create real engagement, asynchronous models (as in Scramble) are far better, because people love to email their friends and call them back to the game.

Some other interesting notes we picked up during the course of the event:

• Facebook advertising is starting to get very profitable. Peanut Labs announced that they distributed $200K to their members just in the past 1 week.
• New advertising models (besides CPC and CPM) are emerging inside the Facebook Platform. The CPI (Cost Per Incentive) model of SocialMedia and the CPE (Cost Per Engagement) model of VideoEgg are just a couple of examples. With CPI, the user is invited to install other apps; with CPE, the user is shown a lightbox page which appears as a new layer on top the Facebook canvas page that he was actually looking at.
• Facebook called on people to produce productivity apps and focus on the long-term value, not viral growth.
• The OpenSocial crew presented Shindig, an open source, Apache-incubated OpenSocial framework for those who are interested in hosting OpenSocial apps inside their network. They have also introduced CAJA, a Javascript sanitization sandbox which allows Javascript code to be embedded into OpenSocial apps safely.
• Investors noted that app makers shouldn't expect to become instant millionaires with their applications. They also highlighted the importance of long-term value and stickiness.
• Facebook app makers are concerned about Facebook's moves to release their own apps; like Pages. They think that Facebook should just provide the platform and not be involved with new apps anymore.

A Brief Interview with Jia Shen

At the end of the day, I had the opportunity to sit down with Jia Shen, the CTO and co-founder of RockYou! and ask him a few questions. Jia acknowledged to me that his company has had a tremendous advantage over others by entering the field early, he also said that these platforms are becoming less permissive for viral growth because of their new safety restrictions - which is why those who plan to make a Facebook app are strongly encouraged act as quickly as possible.

I asked Jia about their ad network, how it was going, and what is the percentage of revenues from non-RockYou! inventory. He told me that it is going well and that the non-RockYou! inventory now accounts for a majority of the revenues generated on the network. He noted some of their clients include Coca-Cola, Johnson & Johnson, and Sony. He also said they are not planning to enter the long-tail advertising race right now.

As for localization and the growing number of international social networks opening up new platforms; he said that they will look at the specific characters of the network that they are entering and make necessary changes to their apps if required. But he added most of their apps are generic so they don't expect big changes. I asked him whether they're planning to make acquisitions to get the leading app makers in outside markets, and he told this is not their primary strategy yet, but if they do, their first target will be big demographics such as apps for Spanish speaking countries.

Jia also said that they have no preference between the Facebook Platform and OpenSocial, they will embrace both, but they support standards and the idea of writing once, running everywhere.

Conclusion

All in all, GSP was a very useful event for anyone working in the social networking industry. There were a number of enlightening sessions and key people who drive the social future of the web were in attendance. As Charlene Li noted, users will get suffocated without the portable social graph; so as a web developer, you'd better catch up with these latest developments and get ready for the future of the web in order to keep your visitors happy. The move of Sheryl Sandberg from Google to Facebook is more than enough to tell you where we are headed.