Inside Facebook

Two weeks ago, Facebook’s Dave Morin said to look out for big news on the data/identity portability front at Web 2.0 Expo SF.  Today, Facebook announced Facebook Connect, a comprehensive set of Platform services designed to let users take their Facebook identity, friends, and privacy anywhere on the web.

The idea behind Facebook Connect is a big one. As owners of a very large and relatively authentic part of the social graph, Facebook wants to allow users to share the identity, privacy settings, and friend lists that they have established on Facebook to application providers around the web. Historically, most social apps have attempted to build their own social graph (i.e. many failed social networks) or have punted on the idea of identity altogether (i.e. blog comments). With Facebook Connect, users can trust that their privacy will not be violated when they share their Facebook identity and friends with other sites.

Facebook Connect has 4 main features:

• Trusted Authentication. Facebook users will be in total control of permissions granted (though the identity crowd will note that it’s a proprietary authentication system).
• Real Identity. Users can bring their real identity with them wherever they go on the web, including basic profile information, profile picture, name, friends, photos, events, groups, and more.
• Friends Access. Users will be able to take their friends with them wherever they go on the web. This will allow developers to create an entirely new class of applications leveraging much deeper social context.
• Dynamic Privacy. Users’ Facebook privacy settings will follow them around the open web. For example, if you change a profile picture or change a setting, this will automatically be updated in Facebook Connect partner websites.

With Facebook Connect, Facebook hopes to take the Facebook Platform to a much deeper level. If Facebook is able to establish itself as the owner and steward of identity, privacy, and friends across the web, the company’s enterprise value will increase significantly.

As most people in the Facebook developer community are aware, the upcoming Facebook profile redesign is the biggest and most far-reaching project at the company so far this year. Its impact on users and developers will be significant, and Facebook hopes it will create a better and more valuable experience for all.

Today, Facebook is announcing detailed specs on how Facebook applications will interact with the new profile page. On the whole, the upcoming updates to the profile page will have a big impact on how users interact with each other and express themselves with applications. In order to get a detailed look at all of these changes, their significance for developers, and Facebook’s hopes for their impact on user experience, I sat down with Ruchi Sanghvi and Josh Elman, Product Managers of the Facebook Platform and leads on the profile redesign effort. Here are all the details.

Developer Documentation & Sandbox

Starting today, developers will have access to documentation on the upcoming profile changes on the Developer Wiki (see links below). In order to create a level playing field, everyone will have the next couple of weeks to read and absorb the changes and think about how they will affect their applications. Starting in late May, Facebook will open a developers’ sandbox that will allow developers to test their apps. A few weeks after the developers’ sandbox opens, Facebook will push the new profile design live to all users. If developers do nothing, their apps will continue to work as is. However, in order to take advantage of new opportunities for integration with the redesigned profile page, developers will need to update their apps in several ways.

With the profile redesign, Facebook wants to give users more control over their profile and make the profile more communication focused. In order to integrate with the new profile design, developers are going to need to adapt to and take advantage of several new integration points on the Facebook profile:

Changes for App Profile Boxes

Application profile boxes will largely be moving to a new home in the updated profile design. The new Facebook profile is getting 5 default tabs: Feed, Info, Wall, Photo, and Boxes. All current profile boxes will be moved to the “Boxes” tab with the update (the title of the “Boxes” tab is still up in the air). The specific migration flow has not yet been determined, but the specifics of how Facebook implements it and helps users decide how to organize their new profile pages will have a big impact on developers. On the Boxes tab, users will be able to add, remove, and organize their profile boxes.

In addition, the Feed, Info, and Wall tabs will have a left-hand column that will hold “main profile” boxes that will be height limited (to about 250 pixels) but otherwise will be very similar to current profile boxes. Users will be able to move their profile boxes between the Boxes tab and the “main profile” area fairly easily. Finally, users will be able to have up to 5 boxes in their “main profile” area - more than that will be put into the “Extended Profile” that developers have come to know well since the initial profile cleanup tool.

The New Info Tab

One of the new integration points for developers in the profile redesign is the new Info tab. This will be a very structured tab, much like the Info sections on the current Facebook profile page, containing key-value pairs on which information will exist in comma separated text or thumbnail lists. Unlike the Info sections currently on the profile page, content added to the Info tab from apps will link directly to the application (not Facebook search results). There will be an in-line editing flow on the Info tab for users; Facebook will provide a suggested type-ahead editing interface.

Users will be able to add content to the Info tab from within application canvas pages (in FBML). After users see a popup confirmation, the content will be added automatically. Each application will get its own section on the Info tab that look much like the different Info sections currently in Facebook’s current profile.

New Application Tabs

Another major new integration point for developers is app tabs. These will be built like “semi-canvas” pages, meaning Facebook will be proxying the images but hitting the apps for the tab’s content. Like profile boxes, app tabs can’t be built with iframes or using auto-play Flash, but developers will be able to build fully interactive FBML pages.

The only place users will be able to add tabs is on the profile itself, using the “+” button to the right of their default tabs, and up to 6 tabs will appear (the rest will be accessible via a “More” dropdown). Facebook will provide a drop-down list of users’ applications, sorted by those they use most. Facebook’s intention for app tabs is that they’ll meet the personal expression needs of those users for whom profile boxes are not enough. However, the page is also a place visitors will be able to initiate interactions, like starting a game. Developers won’t be able to create different views for the profile owner vs. profile visitor - similar limitations to the current app box FBML rules to prevent profile page spam.

Mini Feed 2.0

The focal point of the profile redesign is the new Feed tab. For profile owners and friends, it will be the default tab you see. For profile visitors who aren’t friends, the Info tab will be shown by default (Facebook says this is because it’s more likely you’ll want to learn basic information about non-friends, like where they live or go to college, before you want to learn detailed info about their recent activity).

With the new Feed, there will now be 3 different feed sizes:

• One-Line Stories. This is very similar to what currently exists on the Facebook profile. In the profile redesign, app developers will still be able to use the API to publish feed stories to users’ profile pages.
• Short Stories. Short stories will be slightly larger feed stories, that will have limited FBML. Users can embed images and (soon) Flash.
• Full Stories. These will be up to 700 pixels of free form FBML.

Unlike one-line stories, app developers will not be able to automatically publish short and full stories via the Facebook API. Instead, users will have to approve them. Developers can use feed forms to create multiple versions of feed stories that the user can choose from for publication. When developers want to publish a short or full feed story, users will be given a popup preview of the story, the choice of size (one line, short, full), and the button to approve it. This is a major shift from the current framework of auto-publishing feed stories, but Facebook believes it must give users more control over their profile in this way in order to keep the Feed from becoming spammy.

After a story has been published, users can change which version (one line, short, full) of the feed story is published in their Feed via UI in the feed itself. This means that app developers can publish one line stories automatically, and users can “make them bigger” later on. Facebook will encourage developers to create feed stories in all sizes, so that users can choose which one they like best. Obviously, larger stories are likely to generate more application click traffic than smaller stories.

In addition to being able to publish stories for app users themselves, developers will also have the power to let users publish stories into their friends’ feeds (for example, giving a gift). Once the sender approves the feed story, it will automatically appear in their friends’ feed. However, recipients will be able to change or remove the story later on.

Since removing passive feed publishing earlier this year, developers have been unable to publish feed stories for users not using the app. Now, users will again be able to share app content with their friends not using the app (currently or at all).

The New Publisher

The new publisher flow (a significant upgrade from Wall attachments) will allow users to share content from apps both on their own and friends’ profiles. It’s the main place Facebook wants profile viewers and owners to interact, and this is the change that’s likely to have the most impact on users.

At the top of everyone’s feed will be a way for you to create content on their feed or yours. For example, you may want to add a photo or video, write a note, send a gift or song, draw graffiti, or share links. The interaction will be similar to the way wall attachments currently work. (Facebook thinks of its Wall as just another application that integrates into the Feed’s publisher flow.) When you create content, it will look like a feed story and go directly into the feed. The default feed size will be long for your own profile, short for your friends’.

When you want to attach content to a feed, you’ll be shown Facebook’s default apps (like photos and notes), as well as apps that users have recently used to create content on their feed or otherwise used pretty recently. (In addition, when visiting others’ profiles, users will be invited to use apps that others have recently used to create content on their profile.)

Developers will be able to design the publisher interface that users will use to share content from their apps. For example, users may want to be able to search for songs. And in addition to the “attach-post” flow, developers will have the opportunity to create a multi-step wizard interaction (though like profile boxes, they can’t be iframes, but can be FBML or on-click Flash). Finally, developers will have the option to display a text comment field (or the app can provide that themselves).

Finally, developers will have option to specify different publisher flows for your own vs others’ profile if they want. For example, adding a video would be pretty similar, but it would be weird to send a gift to yourself.

In Summary

The Facebook profile redesign will have a big impact on the developer community. Developers will need to think long and hard about how to make their apps more valuable to users, specifically in the integration points that Facebook is providing. While much of the first year of the Facebook Platform has been about virality, it will be harder for purely “viral” apps that don’t provide much value to users to grow - the second year of the Facebook Platform will be about engagement.

Kicking profile boxes off the default view of the profile page will definitely make it harder for users to discover and re-engage with applications. Facebook must embark on a major user education campaign to help users understand the new tools they have to organize application content on their profile page (specifically, moving boxes between tabs and adding new app tabs) and share application content with their friends (specifically, the new feed publisher).

Ultimately, Facebook really doesn’t want the profile redesign to screw developers. They’re trying to find the right balance between user and developer interests to allow everyone to have a healthy long term relationship. If you have thoughts on what these changes will mean for your applications, now would be a good time to send your thoughts to the Facebook Platform team at developer-feedback@facebook.com. The developers’ sandbox will be opening shortly.

Developer Wiki Resources

• New Design Integration Guide
• New Design Platform Changes
• Details on Publisher
• Details on Narrow Boxes
• Details on Application Info sections
• Details on Feed Wall
• Details on Profile Tabs

Developers can find the official Facebook announcement here.

Facebook developer Jesse Farmer has slurped the Developer Forums for the last 3 months and found some powerful trends: since January, posts per day are down 51%, and and the number of contributing developers is down 27%. In addition, Farmer has tracked daily active user levels for apps launched during that same time period, and the numbers correlate: the later apps have been launched, the lower daily activity levels have been. Farmer concludes that it’s getting harder to have success on the Facebook Platform, and developers are turning their interests elsewhere.

While it’s clear from Farmer’s numbers that developer activity peaked in January, the most important signals of the Facebook Platform economy are ultimately not forum posts but rather financials. As the initial wave of interest in the platform settles out and those working on cracking the monetization nut march on, the true test of the viability of the Facebook Platform is whether a substantial number of entrepreneurs will be able to establish successful businesses that can weather ongoing changes in platform conditions.

The good news for developers is that as social networks compete for developer attention, they will have to prove that the economic health of their developer community is strong. That game is still very early.

The Facebook Profile Preview team tonight unveiled another screenshot of the upcoming Facebook profile page redesign that shows the new “Boxes” tab for the first time. While the main Feed tab will only have room for 3 app boxes by default, the Boxes tab could potentially house dozens of application profile boxes.

According to Facebook, users will be able to choose which boxes appear on this page directly from the tab itself, and users can move boxes around just like they currently can on the profile. Notably, the two columns on this page leave room for wide-format profile boxes, previously thought to be getting the boot from the profile altogether.

Facebook says that, “Our intention is that the ‘Boxes’ tab can be a good place for your friends to go to learn more about you and your history on Facebook, in addition to learning what sort of applications you like.”

Clearly, Facebook faces tough design decisions here as it attempts to balance and align user and developer interests. Without the “Boxes” tab, many developers were concerned that their profile boxes would disappear altogether when Facebook migrates users to the new profile page design. Do you think that this approach is a good way to strike a balance?

While client libraries for Facebook Platform developers exist in many programming languages, you can scratch official support for Java off the list. Facebook announced today that as of next Tuesday, it will no longer be supporting its Java client library, instead relying on community-created libraries.  As Facebook’s James Leszczenski writes,

We encourage all developers who are interested in continuing to develop in Java to consider some of the open source alternative client libraries listed on the Wiki here. While the official Java library should have no immediate problems with continued use, we nevertheless recommend that you use a client library that is kept up to date, in order to best take advantage of any new functionality that is added in the future.

While the number of apps written in Java is small, this is likely to be an inconvenience to those that have chosen to build using Facebook’s official libraries.

While much attention has been paid to the vast volume of inventory that’s been created by social networks, remnant-style CPMs that most social networks earn is not sufficient to sustain a real business for most developers. When the dust settles, can apps hope to do any better?

Andrew Chen has noted 5 reasons why social networks are hard to monetize. While developers face many of the same monetization challenges, developers have two big advantages over the social networks that can actually enable them to achieve higher CPMs than the social networks themselves:

1. App usage is a better signal for intent than profile data.

While users who indicate an interest on their profile may not have significant purchase intent at any given time, users actively engaged in applications around that same interest are more likely to conduct a relevant transaction. Of course, your app needs to serve a need in high value verticals like travel, media, finance, or shopping.

For example, it’s hard to know with much precision when users who list “skiing” as an interest on their profile page are most likely to purchase a lift ticket just based on profile data.  While this is great for basic targeting, it’s nowhere near the level of intent signaled by a Google search. However, users actively engaged in an application used to research current ski conditions and find friends going skiing this weekend are much more likely to have monetizable intent. Whereas profile data is often write-once-live-forever, application usage is a better indication of activated interest in a given vertical.

2. Apps offer better opportunities for brand alignment and integration.

While Facebook offers Pages and Social Ads for brand owners interested in building a presence in Facebook’s social environment, applications can go much deeper and offer experiences much more directly aligned with the values of the brand. With Facebook Pages, brands can only achieve fairly generic levels of user interaction. On the Page itself, users can post messages, upload photos and videos, and sign up to receive future updates. When users become a “Fan” of your brand, their friends will see a message that says, “Justin became a fan of Colgate Toothpaste.”

All of these things are great, but they leave a lot of value to brands on the table. With apps, brands can create more directly aligned and integrated brand experiences - both within the application itself AND in the “viral” messages that users send to their friends.

For example, Federated Media recently did a BMW “What Drives You?” campaign with application developer Graffiti in which users created over 6,000 entries and shared their creations with their friends. Which is a more valuable brand experience to BMW - looking at a Page with a logo and messaging, or drawing the car of your dreams? Which is a better feed item to be seen by millions of friends - “Justin became a fan of BMW” or “Justin challenges you to show What Drives You in the Graffiti Car Contest”? You decide.

It’s not all hunky-dory

To be sure, app developers still face many of the same challenges that social networks do. There’s a lot of inventory, and not all of it can be valued eqaully. Brands are concerned about how they will appear in close proximity to user generated content. And - as always - apps have to play by the rules the social networks themselves set, which could change at any time.

At the end of the day, big opportunities still exist for app developers to monetize well on social networks. Developers need to pick verticals that capture high-value intent/interest and partner with brands to create deep, integrated experiences. Picking the right space? That’s up to you. How to partner with brands? We’ll have more on that soon.

Toronto-based Refresh Partners has been providing demographic analytics tools for its clients to analyze their Facebook apps’ users for months. Now, they’re making that service available to everyone through Refresh Analytics.

Refresh Analytics provides developers with daily snapshots of geographic, age, gender, education, political, and religious distribution, and allows developers to see trends in its user base over time. However, it does not provide behavioral information like page views or time spent.

Kevin Lister, founder of WedSnap, said that his service has been using the service throughout the beta program. “The analysis of demographic information has really helped us go after the right advertisers”, he said.

Interested developers can check out the Refresh Facebook app.

Tonight, Facebook changed the viral channel limits for application developers, making it harder for spammy app developers to keep growing and rewarding developers with good user feedback ratings with more opportunity for growth.

The changes apply to requests (invitations) and notifications. Depending on user feedback ratings on your apps’ requests and notifications, applications sending high-quality messages through these viral channels will experience higher limits, while apps that receive bad ratings will get lower limits. As the Facebook Platform team writes,

Notifications now have more buckets for greater granularity and a broader distribution of thresholds, so applications in higher buckets should receive larger allocations that before and lower buckets should receive smaller allocations. Requests still has the same number of buckets, but the distribution has also been pushed wider, resulting in more requests for some applications and fewer requests for other applications.

Facebook’s continued optimization of its viral channel throttling system is a good thing for the overall health of the Platform - as long as it still allows new applications the chance to bubble up.

As most Facebook Platform developers are aware, a major ongoing project at Facebook right now is the profile redesign. While the redesigned profile page will not be released to developers or users for several more weeks, Facebook has released an updated concept mock for the new profile page tabs that split the previously combined Feed/Wall tab into separate tabs.

As Facebook’s Profile Preview team writes, the intentions for the tabs in this version are as follows:

• Feed: Previously called the ‘Wall’ tab, this tab has a combination of your Mini-Feed stories, as well as your wall posts.
• Info: Previously called the ‘About’ tab, this tab has more static content, such as your ‘About Me’ section and personal info.
• Wall: This tab is new, and is simply your Wall as you currently know it. Only wall posts show up here.
• Photos: This hasn’t changed. This is just a tab that consists of your tagged photos as well as your photo albums.

What does this mean for developers? Not a lot just yet. While designs for the updated profile page are not yet finalized, more tabs for Facebook features would generally mean less room for apps.

Recently, I spoke with Jeremy Liew of Lightspeed Venture Partners about the future of gaming on social networks. Jeremy has been writing frequently about social games on his blog and recently moderated a panel on the topic at Web 2.0 Expo.

During our conversation, Jeremy brought up some important points regarding the way the core mechanics of social games are designed, specifically regarding the different ways that some game developers are integrating “viral channels” on social networks into their games:

It’s not an accident that the biggest games on Facebook are not owned by SGN or Zynga: Scrabulous and Friends For Sale have been growing independently.  They’ve been baking virality directly into the game mechanic, not layering it on top. They’re developing game mechanics that are native to social networking environment - not just using social networks as a distribution channel. This has only been going on for a few months, and so it’s clear to me that there’s still a lot more opportunity to optimize that.

Read the full interview at Inside Social Games