Inside Facebook

Recently, I was selected to be a judge in Bebo’s B.E.S.T. Developer contest.  Over 2,000 applicants participated, and after 2 months, 18 were selected by Bebo to be judged.  Frank Gruber, Dave McClure, and I (Jesse Stay) all had the opportunity of trying out each one and judge, based on various criteria, what we thought of them.  The results were all tallied up by Bebo, and the top 5 were announced at an awards ceremony.  Those included Kickmania, Banana Grams, Robokill Trainer, and Solid Sudoku. The winner, Bowling Buddies, by Playfish, deservedly won Grand Prize due to it being a very engaging and visually appealing app with functional UI.

As a social app developer, and author of O’Reilly’s FBML Essentials, I’ve been through these contests before as a participant, often with a blind eye towards what the judges would look for and how I could get their attention.  After all, in this particular contest, over 2,000 applications were competing for the top prize, but in many Facebook competitions there are 10 times that.  You’re a needle in a haystack to the judges.  I thought, after experiencing this and a few other judging opportunities, that I’d share a little on what catches our, as judges, attention:

1. Functionality - I’m starting with the essentials, and they were very important in the judging process for this particular contest.  Functionality was key.  Believe it or not, even some of the top 18 finalists that we judged still had major flaws, bugs, and issues with functionality. Some apps took several tries to get working after install. Others had bugs within the app itself. I understand you didn’t have much time to write it, but neither did any other developer in the contest, and time is not on the Judges’ side either. That brings me to #2.

2. K.I.S.S. - Keep it Simple, Stupid! There were a few apps that didn’t even have About pages. As a Judge, this is crucial. Remember, some of the judges (those at Bebo, or Facebook, or whatever the host of the contest is) are going through thousands of apps. Others, like me, are going through just 18, but even 18 is a lot to go through on an already very busy schedule. (They don’t call Dave McClure “Master of 500 Hats” because he sits around and judges contests all day.) Your About page is your cover letter - if the Judge never even tries your app, what do you want them to know about it?

Keep your apps very simple, but fun, engaging, and useful. The more you can fit into a small package without being overwhelming, the better. Make it as easy as possible for the judges to get through most of your app, very quickly, and easily.

3. Visual Appeal - Those apps that were a little more appealing visually caught my attention more than those that didn’t. As an app developer myself this doesn’t matter as much to me, but I did find myself sticking around on apps I could “feel good” in a little longer than those that didn’t spend as much time on visual appeal. It made me feel like the developers cared a little more if they invested in a graphic designer to design their application.

4. Engagement - I highly recommend you try out the 5 apps above.  I rated I think every one of them pretty high, and the Grand Prize winner was definitely my favorite.  Bowling Buddies drew me in - in fact, I wrote on Twitter at the time I was playing it, “The good apps in this contest really suck you in.” I liked it so much and was having so much fun playing it I wanted to share it with others and even brought my 6 year old son over to play it with me. I really wanted to tell others about it.  Spend some time having your friends and family play your app - do they share it with their friends (and not because you sent it to them)? How long do they spend playing it?  Think about the types of things your buddies at work send to you and would waste work hours on (seriously). Engagement is very important.

5. Virality - I was very surprised that some of the apps we judged did not make it very easy for me to share with my friends.  Virality is key, and central to the success of any social application. Your application should sell itself.  Your users are your new sales and marketing staff, and best of all, they don’t cost you anything! Make your app engaging enough to make me want to share, and when I want to share, make it as easy as possible, at as many points in the app (without being spammy) for me to share it with my friends. A good judge of social applications will notice this.

6. Get to know the judges - This isn’t an exact science.  The three judges for Bebo’s contest were announced ahead of time.  Our names are all over the place, and we’re all on Twitter, Facebook, and Bebo.  It’s very easy to learn about us, what types of things we like, the quirks we hate in applications, and the types of innovation we’re looking for.  Take advantage of that to get to know the judges ahead of time.  I don’t think a single app developer tried to chat with me and get to know me during this contest. While I probably would not have let it influence me, it could have given them some useful information on what I’m looking for.  This isn’t the case for all contests, so tread lightly (it could backfire), but at the very least, get to know who will be judging your app.

As for me? I was looking for the most useful apps that could be viral and engaging. I like things that change the world, things I haven’t seen before but catch my eye. Surprisingly not many I judged did that, and perhaps that was rightly so since Bebo does target more of an entertainment demographic. With the short timeframe it’s also hard to do that, which I understand. Perhaps yours could have been one of those?

7. Read directions - In this contest, Bebo laid out their instructions clearly.  I don’t know if they turned any apps down due to not following directions, but I bet they did, and you would never have heard of it. Read all the contest rules ahead of time, read them twice, and be sure your application follows them, to a tee, when you submit your application to the contest.  If it does make it through, the judges will notice, and then you’re just wasting the judges’ time at that point.

8. Be original! - Remember that as judges, we’re looking at many applications.  Therefore, those that stand out will be those that haven’t been done before.  I was very impressed, and amused at the Kickmania app, due to the originality of it all, and how at the same time it sucked me in and was very viral. While perhaps not something that would change the world, it was very original in the idea, something I had never seen before, yet wanted to play over and over again. Surprisingly, Bowling Buddies was the same way - while Bowling itself isn’t very original, the way they crafted the Bowling game and the way you play it I thought was very original.

This was a fun contest to judge, and I learned a lot from it. How you craft your application and the amount of time you spend paying attention to these details will make or break your application in these contests.  If you can follow these 8 steps, your chances of being noticed will go up 10-fold.  Congrats again to the winners!

The London Facebook Developer Garage is coming up on Wednesday 8th October at Sun Microsystems, 45 King William Street, London. Sun are generously sponsoring the event again and are also providing free beer and pizza.

As well as the usual networking opportunities there’s a busy line-up of speakers:

• Phil Clandillon and Steve Milbourne from Sony BMG talk about an application based competition they ran across Facebook and Bebo to promote Michael Jackson, lessons they’ve learnt, and what else they’re doing on the platforms.
• Chris Thorpe from MySpace discusses the changover from Open Social 0.7 to 0.8, and why sometimes it still makes sense to use Javascript in normal applications.
• I’ll be talking about Facebook Connect with a technical overview and a practical view of pros and cons.
• Mat from Sun talks about what they’ve been doing with Zembly (a tool to build and host applications across multiple social platforms)
• Toby from Nudge shows us their latest application, Spotted!
• Iskandar from Nudge goes over the latest developments on the Facebook platform.

The event starts at 6:30pm and tickets are available from Eventbrite or at the door.

Friendster, the biggest social network in Asia, will announce later this evening that it is expanding its application developer program by adding support for the Facebook Platform API. With tonight’s launch, Friendster will be the first social network to have implemented support for both the OpenSocial and Facebook Platform application development standards. Friendster launched OpenSocial support just this August, after implementing its own application platform in 2007.

The Friendster launch is a big opportunity for Facebook application developers interested in Asian markets. Facebook developers will be able to easily port (most parts of) their Facebook applications over to Friendster with very little work, gaining access to Friendster’s audience of 57 million users. According to comScore, Friendster’s user base only has a 22% overlap with Facebook’s, so Friendster offers access to an audience largely untapped by most Facebook developers.

In addition, Friendster is less restrictive about which parts of the application can be monetized. Developers can put ads on any real estate they control, including the profile page.

Facebook Platform API categories that will be supported as of launch tonight include: users, friends, notifications, photos, FBML, FQL, authentication, requests, and profile. The news feed API will not be supported, but it will be in the future. Friendster’s Director of Marketing Jeff Roberto says that Friendster intends to keep its Facebook Platform implementation up to date with Facebook’s, but it will just take a little more time to initially get in sync.

“Our vision with the developer program is to build a very open platform and to make it easy for developers who have built on common standards like Facebook and OpenSocial to deploy on Friendster,” Roberto says. “We will maintain our Friendster APIs, Facebook APIs, and OpenSocial APIs.”

Friendster becomes the second major social network to license the Facebook Platform API under the Facebook Open Platform (fbOpen) initiative. Late last year, Bebo launched its platform using Facebook APIs, shortly before being acquired by AOL.

comScore released its June 2008 social networking numbers last week. After taking a little bit of a deeper look at social networking over the past 6 months, I think there are some really interesting trends here:

• Overall, Social Networking total audience grew 64 million (12%) during the last 6 months. (That’s over twice as fast as the total internet audience.)
• Of that, Facebook and hi5 are winning the vast majority of new users coming into the category. hi5 was by far the fastest growing site at nearly 80%, or 25 million uniques, during the period. Facebook grew by the most absolute uniques at 34 million, or 35%. Together, Facebook and hi5 grew by 59 million uniques, accounting for 67% of the growth of the top 6 social networking sites during the first half of 2008 (and 92% of the total category’s net growth during the first half of 2008).
• The next closest in unique terms was MySpace at 10 million, or 10% growth during the period. However, Friendster (7 million uniques/25%), Orkut (9 million uniques/35%), and Bebo (2.7 million uniques/13%) grew at faster rates.
• Facebook is clearly the largest social network in the world by total uniques, at 132 million. MySpace has saturated the US market and is not growing as fast as Facebook and hi5 internationally, where most of the growth is coming from.
• For those that don’t yet know, hi5 is exploding. hi5’s massive 80% growth over the last quarter now firmly entrenches it as the #3 social network in the world. hi5 is the leading social network in 25 countries across Latin America, Europe, Asia and Africa.
• During the first half of 2008, the market became increasingly concentrated amongst the 6 market leaders. Although uniques to the leading 6 social networking sites increased by 88 million, uniques to the other 276 social networking sites in the category actually fell by 24 million from December to June. Despite all the new social networks being launched, it’s proving to be really hard to gain traction at this point.

Update: I think the calculations involving the “Other” bucket are incorrect: the Unique Visitors metric is not additive. Also, the data refers to individual months, not six months at a time.

While Facebook is rumored to be launching Facebook Connect, a new way for third party websites to connect with users’ Facebook profile and friend information without leaving the site, the Bebo Platform team have launched some interesting API’s of their own: today, Bebo announced new external authorization APIs which will allow third party web and desktop applications to access your Bebo account.

Developers interested in the Bebo external auth APIs can find more documentation here.

This is a guest post by Blake Commagere, creator of the “monsters” series of applications on Facebook, Bebo, and other social networks.

How much work is it?

Several Facebook developers have still not ported their apps to Bebo despite the ease with which it can be done. The team at Bebo has gone to great pains to ensure that porting your app from Facebook to Bebo is very, very little work. Given that you can access Bebo’s audience of 80 million users by spending just a little time on it, you should begin your port immediately - my applications (Zombies, Vampires, Werewolves & Slayers) took me about 9 hours total to port over. Here’s a quick guide to help you get started:

Two very important decisions to make:

1. Whether you allow cross-platform functionality.
2. Whether you use the same code base.

Allowing cross-platform functionality

Cross-platform functionality is simply any set of features that allow users on Facebook to interact with users on Bebo. In general, this is not a good decision because it is a non-trivial amount of work and will add value to your application in very, very few instances.

1. While the average Internet user may have accounts on a few social networks, they are only active on one.
   • Keep in mind that engineers are an exception to this rule, but most users are NOT engineers
   • Very few active users on both networks means very few that would need a seamless experience between the two networks.
2. Transferring a user’s settings from one network to another would require authenticating them against both social networks.
   • This would either have to be done by Facebook and Bebo through some sort of collaboration
   • Or, you will have to authenticate the user by prompting for them to log into one social network from an experience in the other - a very confusing user experience.
3. Any tables indexed by user_id will have to be indexed by social_network and user_id to prevent collisions.
4. Friend graphs do not map across the networks - i.e., your Facebook friends do not directly correlate to your Bebo friends.

There are cases where cross platform functionality does make sense and is not much work. If your app creates a synchronous experience with strangers - for example a flash poker game - then the user can be presented with a room that allows them to play poker against members of many different social networks and creates a larger community. However, functionality that allowed you to transfer your poker history and settings from Facebook to Bebo is a feature that would likely get very little use and add a non-trivial amount of work to your porting effort.

Using the same code base

Bebo has gone to great lengths to make using your same code base for both Bebo and Facebook very easy. In general, this is a very good decision for the following reasons:

• Your feature set will always remain in synch
• Bug fixes only have to happen in one place

If you are using MVC architecture, then you should be able to reuse your models and controllers with very little adjustment. As Bebo does not yet have FBJS support, the view portion of your application will require some work if you use FBJS.

If your app is not using an MVC architecture and you extensively use FBJS, then your code will have several places where you have to case out the different social networks and it may be worth branching your code base rather than reusing it.

Recommended Steps

(assumes you want to use the same code base and are not allowing cross-platform functionality)

1. Ensure you are using 64 bit integers for your user ids (e.g. in MySQL bigint rather than int).
2. Create a top level configuration file that determines whether the connection is coming from Facebook or Bebo and sets a parameter or define establishing which network has connected to you.

   * You can do something like this:

* if ($_REQUEST['fb_sig_network'] == 'Bebo')

define('CURRENT_SOCIAL_NETWORK', 'Bebo');else

define('CURRENT_SOCIAL_NETWORK', 'Facebook');

3. Merge the client APIs for Facebook and Bebo. You can start with the Facebook API and add cases that handle connecting to either Facebook servers or Bebo servers based on the parameter you established in Step 1. (Note: I am hoping to open source a merged client API for Facebook and Bebo in the very near future - this will make this step unnecessary!)
4. Handle any cases where you relied upon aspects of the Facebook API that are not yet supported by Bebo. These API calls aren’t supported by Bebo yet:
   • AuthCreate/getsession (but this is coming very soon)
   • Admin.getAppProperties/Set AppProperties
   • Batch:run
   • Users.hasAppPermission
   • Users.getstatus
   • Friends.getlist
   • Datastorage
   • Marketplace
5. If using FBML, handle any cases where you use facebook tags not yet supported by Bebo. Keep in mind: FBML is supported by Bebo - it SNML has the same syntax, but you do NOT have to switch from using to. Bebo allows you to use either! Currently, here are the tags not yet supported by Bebo:
   • mobile
   • require log-in attribute
   • page-edit-admin-header
   • visible to connection
   • attachment preview
   • if is-in network tag
6. If you want to build functionality supported by Bebo but not Facebook, handle those cases properly. Here are the API components on Bebo that do not exist on Facebook:
   • Bands
   • Playlists
   • Songs of Bands
7. Keep in mind that Bebo does not yet support FBJS (SNJS), but that will be coming very very soon! Additionally, pre-load SNQL launches the week of 6/02/08, so use it!
8. Abstract out Facebook specific links or references. For example, anything specified here: http://developers.facebook.com/documentation.php?v=1.0&doc=other will most likely not be as trivial as replacing the word “facebook” with “bebo” in these urls.
9. Release on Bebo and enjoy!

About the Author

Blake was a Founder at Mogad.com, which later became Social.im. Blake is also the creator of the superviral Zombies, Vampires, and Werewolves games on Facebook. Prior, he led the development of the Causes on Facebook application. Blake was also an early engineer on Plaxo’s client team and a founding engineer at BuildForge (acquired by IBM).

We’re here at the 1:10pm session and this panel is covering platform options. The panel:

• Oren Michels, CEO of Mashery
• Jessica Alter, Director of Platform and Business Development at Bebo
• Daniel Burton, Senior Vice President, Global Public Policy for Salesforce.com
• David Jones, VP, Global Marketing at Friendster

Oren: What are you doing with your platform? What makes it a little different? Whats new with the platform since the last GSP in San Diego?

David: We are the seventh largest website of any kind in the world. We are the third largest social network in terms of traffic. We are the number ones in terms of engagement relative to other social networks. 38% of the entire internet population is in Asia and Friendster is the largest social network in that region. We have 350 applications live on the platform today with 1000 currently under development. The revenue model is completely open on our platform. In the last few months we’ve added some viral channels for every app and every developer.

Jessica: We are a social network focused on community, entertainment, and communication combined together. Our users are really engaged in what they’re doing on our site and that is something we are very focused on continuing to grow. Thus far we have over 4,000 applications on our platform. We currently support the Facebook APIs but we do plan to support Open Social in the future. We built a platform because we realize that outside developers could bring to our site many things that we simply couldn’t offer to our users ourselves. After photos, apps are the most popular thing on the site.

Dan: Last fall we announced our platform Force.com. It is the first platform as a service for business. It is a system of tools and application services that allow both developers and corporate IT departments to developer applications and have them run on the Salesforce infrastructure. It is a very powerful environment not only to developer your applications but also to run your applications. The difference between our business platform and the social platforms is the level of security requirements that enterprises really depend on. We created the eco-system with all of the security functionality built in so that developers can put their applications on our platform and have access to an entire universe of users for free.

Oren: It sounds like the monetizing element is baked into the force.com environment. For Bebo and Friendster, are people building businesses on your platform?

David: We are many seeing many developers generating healthy revenue through CPA.

Jessica: Yes, there are a lot of developers making money. Our approach is how can “we” make money, we including both the developers and Bebo itself. We are looking at our rich data and how can we help use the data we have to help developers monetize. We want to offer increased ad relevance for the entire ecosystem. We think that there needs to be some way to share the data we have so that the CPMs developers are seeing are much higher than the .05 or .10 cent range.

Oren: In terms of access to profile info, what are the trends in regards of the balance between privacy and personal data and allowing the developer to take advantage of the rich data that is being collected?

Dan: Salesforce and Force.com is a little different in that regard. We do not see nor collect data. On Salesforce you can buy directly from the developer, so we are really outside of that ad based model. We should all be watching a lot on the public policy side that may alter the behavioral ad targeting environment.

Jessica: This probably is the most delicate balance we have to deal with given the huge opportunity and huge risk. It is a constant conversation that we have internally in the company. We think there are ways to achieve a balance that doesn’t compromise users.

David: First off, there are tons of granular privacy settings on Friendster. In terms of the APIs and the developer program, a developer can only access certain private information once the user actively consents. We’ve put in place controls to keep a tight handle on this issue so that users know whats going on while developers can still build worthwhile applications on the platform.

Jessica: The question we’ve found when talking about lending this information to developers or any third party is do users understand what they’re doing when they give access to their data? The hardest part is educating the users to understand the significance of their actions.

Oren: People definitely don’t know. Inevitably, the bad news is going to happen and policy will quickly enter into the equation. What do you see as the next big thing over the next six months as the era of throwing sheep passes?

David: Friendster is a founding member of Opensocial and we will be deploying the spec in coming months. We are looking for new, compelling, innovative apps. We haven’t seen as many as we would have expected. We are looking forward to some truly next generation apps to make life easier for our users.

Jessica: I’m for the kind of apps that my users are going to use. To me, thats about engagement. What I think is quality is not what my average 21 year old user thinks is quality. We want things that will be really engaging. In the next few weeks we will be coming out with some new engagement metrics and rewarding based on those. We definitely hope to innovate on the monetization side as well.

Dan: We think this is a terribly exciting time considering all of these new platforms. This is the era of “let a thousands flowers bloom.” We really think that what we’re moving towards is the end of software, and you’ve seen that with applications and you’re seeing platforms take that to the next level. The consumer web is setting the pace that the business web tries to interpret and implement in the business environment. The dynamic combination of these two worlds will lend a tremendous amount of excitement to both the business and consumer side.

Q: For salesforce, for developers how much help do you offer to meet the security requirements?

Dan: Not only do you as a developer code in salesforce, but we do also have a security audit. Our team works with developers to make sure that they are following proper security procedures. All of those security issues are automatically handled by the run time environment.

The Bebo Platform team announced over the weekend that it has launched Facebook-compatible support for pre-load SNQL. This basically means that developers can make their apps run a lot faster on the Bebo Platform by specifying beforehand what platform data the application requires instead of making lots of queries as you go.

While we haven’t heard any major news out of the Bebo Platform lately, Bebo has told me they are working on new engagement metrics for applications. These metrics (like time spent, etc.) will reward developers that create the most engaging applications.

It’s been many weeks in the making, but we at Inside Facebook have some exciting news to share today - we are launching a new, experimental second blog called Inside Social Games! Inside Social Games is going to be entirely devoted to tracking the convergence of games and social network platforms.

So what are social games anyway - and why Inside Social Games now? Just as I was surprised two years ago by the way my “non-geeky” friends were using Facebook, I’m finding myself repeatedly being surprised by the way those same people are playing new kinds of “social games” inside Facebook (and MySpace, Bebo, Friendster, Hi5, and Orkut) now. Games that, despite their apparent simplicity or complexity, are being adopted in large numbers by people outside traditional console, PC, and online gaming demographics. Games on social networks are bringing together family and friends like games never before have - and we think that’s pretty darn cool.

Like Inside Facebook two years ago, I’m starting Inside Social Games as an experiment. An experiment into what “social gaming” really means and how social games may or may not be changing the way large numbers of people play and communicate. What the results will be, I’m not yet sure, but I’m excited to see what conversations emerge over the coming weeks and months, especially between the game design and SNAP communities.

I’m proud to announce that I’ve been able to land Chris Holt (also of PC Gamer) to review current and new releases. Chris would love to receive tips on new games, so feel free to contact him at submit AT insidesocialgames DOT com to let him know what you’re up to.

We’re looking forward to tracking not only the games but also the people and companies behind them that are leading the charge into this new and exciting space. We believe some interesting new companies will be created as the worlds of gaming and social networks collide. It should be a fun ride. Please drop us a line any time!

This morning at Web 2.0 Expo San Francisco I moderated the Comparing Social Platforms panel, which was comprised of awesome folks speaking on behalf of the top social platforms on the web today:

• Allen Hurff, SVP of Engineering at MySpace
• Dave Morin, Senior Platform Manager at Facebook
• Jessica Alter, Director of Platform and Business Development at Bebo
• Patrick Chanezon, OpenSocial Evangelist at Google
• David Recordon, Open Platform Lead at Six Apart

While the panel covered a variety topics that each merit more discussion (there are entire conferences devoted to doing just that), this morning the panel discussed key issues for the social platform economy - such as each platform’s philosophy on key product integration points (profile, feeds, viral channels), aligning user and developer interests (engagement, marketing channels), monetization roadmap (ad networks, commerce), data portability, and potential business conflicts (in verticals like photos and music).

Since I was moderating the panel, I didn’t have time to take notes, but here are some paraphrases of some noteworthy panelist comments:

• Jessica Alter: We’re going to be driving developers to focus more and more on engagement. While I don’t have any specific announcements today, things like time spent and page views are going to become more important on the platform.
• Allen Hurff: On the question of verticals like music, of course MySpace is going to continue to work with major music distribution partners around the world. But we really want to remain as agnostic as possible. In fact, I expect there to be many more app feed items than MySpace system feed items.
• Dave Morin: We’re spending a lot of our time right now thinking about data portability, or perhaps it would better be called “privacy portability.” You should have the option to have your privacy settings follow you around the web.
• David Recordon: The term “data portability” almost can’t be used in that way any more, since the branding of the term now means something else.
• Patrick Chanezon: Unlike most of these social platforms, iGoogle is much more of a social content discovery service than social network. It’s a pretty different type of service that’s built on OpenSocial that we think is really interesting.
• Allen Hurff: The MySpace Platform actually represents a pretty big shift just in the last year from the way MySpace has traditionally viewed opening up its data. We’re going to push this even further in 2008.
• Dave Morin: Creating a platform is like creating a marketplace. You have your arbitrageurs, and we have those, you have your hedge funds, they’re like the guys staying up all night finding ways to tweak their way into the viral channels more.
• Patrick Chanezon: OpenSocial is really just an overlap of the most common set of services needed to run social applications. We may begin to see platforms on top of platforms in the coming year.

Despite the early start time (8:30!), by midway through the panel there were folks crowding in the back of the room. We only had time for a couple audience questions before we had to make room for the next panel. Thanks to everyone for a great discussion! For more, check out GSP East.

Update: Here’s a video of the panel, thanks to Simon Chen: