Tonight is the first chance I’ve had to write about the Facebook Developer/ Influencer conference that I went to last week.  The event was an invite-only afternoon of panel discussions hosted by Seth Goldstein of SocialMedia.  Attendees included about 50 developers, entrepreneurs, investors, and a couple of bloggers, and the discussions ranged from “When, if ever, will Facebook start ‘taking back’ core chunks of its platform?” to “What metrics really matter for gauging success on the Facebook platform?”

The conference began at Noon with lunch and networking followed by the first panel discussion at 12:30: “What is Engagement and why is it so important?” In the first session, Dave McClure from 500 Hats and Seth Goldstein  set the stage for the rest of the event.  Dave emphasized the importance of establishing more meaningful metrics for measuring the success of Facebook apps – beyond counting user installs.  His point was well taken… Clearly we need a way of measuring user engagement in apps, especially given that people are often compelled to download apps that their friends send to them and never use them again.  Just because an app has great word of mouth success initially, doesn’t mean that it will latch on for the long haul.  Similarly, time spent on the app isn’t the right measure for success either.  As Seth asked (and I’m paraphrasing), ‘What’s a more important to Facebook’s success- a graffiti app that allows Facebook users to draw for 3 hours, or an application that encourages shorter but more frequent interactions?’

Rumor has it that Facebook will be unearthing at least a few 3rd party app success metrics internally in the next couple of month, but it remains to be seen whether they’ll share this information with the world.  So far, Facebook hasn’t released any helpful metrics for measuring user engagement on apps.  Perhaps this is because they don’t have them, or perhaps they’re holding their cards close to the vest in the hopes of determining the best way to move forward (i.e. by taking back parts of the API they already opened and/or extending new Facebook features that leverage lessons learned by observing user engagement stats on leading apps).  Either way, in order for the 3rd party developer community to flourish on Facebook,  developers will need a better understanding of what makes a winning app and which apps are the most successful based on those metrics.

The next session of the day was about “Creating, Spreading and Scaling Multi Million User Facebook Apps.” The all-developer panel included:

• R. Tyler Ballance from Slide,
  
• Blake Commagere of Vampires / Zombies / Causes,
  
• Dave Genztel from SocialMedia,
  
• Jia Shen from RockYou,
  
• Joe Winterhalter and his colleague, Eric (didn’t catch his last name), of Quizzes and
  
• James Hong from Hot or Not.
  

Slide had over 10 million Facebook users at the time of the conference, and Tyler initiated the conversation by talking about what he thinks users want to do with their friends.  In cases where users have a large number of Facebook ‘friends’ (Tyler had a whopping 630 at the time), Tyler felt it was important to be able to quickly engage with lots of people, while having the option to more intimately engage with closer ‘friends’.   He attributed the success of applications like Happy Hour, which allows users to send virtual drinks to their ‘friends’, to this very phenomenon.  Dave Gentzel from SocialMedia, who developed apps like Happy Hour and Food Fight, agreed that user engagement is key.  In addition, he felt that speed in development was a key component of success.  Anything that’s taken him longer than 2 days to develop hasn’t paid off.

Tyler also brought up the issue of Facebook’s need to communicate more proactively and effectively with Developers.  He and others expressed concerns that Facebook had made some code changes without talking to developers in recent weeks, which resulted in killing thousands of profiles within Slide’s database alone.  Generally, the developer panelists felt that Facebook’s attempts to help a large number of small developers may unintentionally hurt larger developers.  As I’ve mentioned in previous blogs, I’m a huge proponent of investing in developer relations.  When you allow ISVs/ developers (at least large ones) to plug into your API, you should be treat them as strategic alliance partners and give them insight into what you’re planning in exchange for quality assurances.  If you don’t, you run the risk of alienating thousands of your users if/when something goes wrong with the apps that plug into your platform.

Blake Commagere, who helped develop popular facebook apps like Causes went on to talk more about developing Facebook apps.  It took 4 engineers to develop the Causes app (which was written in Ruby on Rails).  Blake pointed out that to develop a successful Facebook app, you don’t need 100 app servers, you just need to make sure the app and database are solid.  By way of example, Causes runs using 11 app servers, which serve 2.5 million users, and it is working well.  Joe and Eric who developed Quizzes, only use 4 servers for their app.  They emphasize the importance of focusing on app quality and investing time in apps that will grow spread quickly virally.

All of the developers mentioned that that Facebook platform is a little sluggish at times.  James Hong from Hot or Not said that to combat delays, his team opted for using Ajax.  The challenge here is that most ad networks don’t currently consider user action as the way by which advertisers pay for ads. Instead, it’s still page changes.  In principle, the ad networks James knows say they’re happy to move towards a user action model, but in the meantime, there are monetary disadvantages to using Ajax on Facebook.  But, most of the panelists seemed okay with the tradeoffs in the short-term because they increase user engagement long-term.  At the time, Hot or Not is apparently making $1000/day off of AdSense, and rumor has it (according to a member of the audience) that Graffiti is making $100,000 month!

Ads were a hot topic on all panels – including the developer panel – with the need for relevant content delivery emerging as a key theme.  Most of the developers on the panel said they’d been approached to do demographic based behavioral targeting of users.  What I found interesting is that the only data anyone would cop to hearing advertisers request is: sex and geography.  If this panel was any indication, for all of those personalization fans out there (of which I’m one), it looks like we’re a ways off from seeing any meaningful movement in this space.

The 2pm panel was on “Facebook Advertising Models.”  Panelists were:

• Aryeh Goldsmith (Acebucks)
• Sourabh Niyogi (Appsaholic)
• Scott Rafer (Lookery)
• Narendra Rocherolle (fbExchange)
• Matt Sanchez (VideoEgg)
• Sundeep Ahuja (Appfuel)

Perhaps the most interesting part of this panel was how so many people are attempting to create ad networks on Facebook.   Matt Sanchez from VideoEgg says he has 20 sales people selling in 4 countries.  Aryeh Goldsmith is creating a new Facebook economy and customer loyalty company by introducing a fake currency to Facebook called Acebucks.  At best, the currency takes off and does just what Linden Dollars are doing for SecondLife.  Appsaholic is hoping to become a fully independent monitor of Facebook Stats (think the AC Nielsen of Facebook).  While, Slide wants to be the “single largest content delivery platform to unique users” across all social media OSes.

How useful are Facebook users to developers and advertisers?  The final session addressed “How to Value Facebook Apps.”  The panelists were:

• Eve Phillips, (Greylock)
• Keith Rabois, (Slide)
• Naval Ravikant, (Hitforge)
• Angela Strange, (Bay Partners’ AppFactory)
• Susan Wu, (Charles River Ventures)

One VC in the audience reckoned that each Facebook user is worth about $500 each.  While, Angela Strange from Bay Partners, whose AppFactory is actively seeking investments in start-ups that make Facebook applications, used a guideline of 25-30 cents per user per app.  In watching this last panel and listening to the excitement relayed through the questions from the crowd and VC responses, it’s clear that there are a lot of people pinning their hopes on Facebook (and other Social Media Platforms) continuing to open their APIs and thus encouraging the growth of a new developer economy that pushes apps for so-called social media operating systems.  There is a frenzy of enthusiasm about the potential, but it remains to be seen whether potential becomes reality.  The VCs aren’t taking any chances.  Each of the panelists said that when they’re looking to invest in a company that develops Facebook apps, they need to see the potential of those applications beyond the confines of Facebook.

For those of you interested in future developments in the Facebook space, Dave McClure mentioned that he’s planning a Facebook conference of his own on October 7-9. Stay tuned to his blog for details.