The launch last week of Facebook's platform initiative, F8, has generated a lot of talk, much of it in the mainstream press. Its a compelling story, Facebook is becoming a platform, out maneuvering Myspace, doing to the web what Microsoft did to the PC. Its a story we have heard before, it seems to recur periodically. However, the announcement last week was mostly about distribution – it didn't involve either deep or open access to Facebook data nor open access to its infrastructure. F8 as it stands today is a partnering platform. This one more small step in a long negotiation that is taking place between web sites on how data is owned, on how its shared between sites and how people navigate through services on one site to another. This conversation is still in its infancy.
XML really began the process of lateral data flows between sites and the vision of the semantic web offers a rich set possibilities — yet it's early days — most sites still operate in vaccum's and most user data is still stuck in proprietary silos. And while the technology certainly needs to evolve so do the scope and kind of business arrangements. The web of contracts, contracts between vertical sites, contacts between sites and users – needs to evolve in order for the vision of the semantic web to reach some of its compelling end points. Weaving, back to the Facebook announcement. What happens next is more interesting than what happened last week. Facebook has taken a different approach to Myspace – who has opt'd to control much of its third party innovation through fairly simplistic interfaces and binary business driven rules, more like a traditional media company, vs. letting the community really build on top of the service in a meaningful manner. As the Facebook platform evolves there are a handful of things I will be watching:
1. How deep are are the API's that Facebook is going to present to the community. Facebook markup language is a proprietary API, the "platform" maybe wide in terms of distribution but its not deep, there is little to no access for third parties to the social data or infrastructure that makes Facebook such an interesting service, and its not open for developers to just build on, everyone accepted into the platform has to be sanctioned by Facebook, the degree of openness, real openness (vs. marketing gibberish) will dictate the depth and the value of the platform. Amazon has done a great job at developing a set of platform services — starting with the affiliate model, extending it into community and then the Mechanical Turk and the elastic computing cloud services. These web services were built step by step along with trust and a degree of openness that surprised many. Pretty much every startup I work with today is using EC2/S3 — if Facebook going to have the same influence over the web application space, if so they need to open up more than a distribution funnel. iLike's weekend server hunt demonstrates a need on the infrastructure side, but the is also a real need re: social data. Offering Facebook users the ability to port social data, their social network across applications and letting applications developers innovate on top of that data set would be really interesting.
2. How will the application metaphor evolve? I see the metaphor Facebook has applied as the most interesting thing in the announcement last week. The web has spawned many interesting platforms for micro application development. Applets, plugin's – from WordPress to Firefox to Myspace there is a an active ecosystem of development around many web sites. But the term application suggests user control beyond a widget or plug-in, applications are often monolithic, the management of applications by the underlying OS is usually benign and in service to the application (get me that device driver) — the term application presents a high bar for Facebook to jump over. To me the use of the term suggests a rich set of API's and a clearly defined layer – a layering of both technical and business terms. Its an exciting challenge to see if they can make this truly an application environments. And if they do, what is Facebook's relationship to these applications? The identity issue below is only scratching the surface of this question. It was fascinating to me that in the announcement last week most of the mainstream press look in the rear view mirror for metaphors — this was going to be like windows was to the PC. I hope not — we don't need another OS, what we need are open development platforms — and open access to data. I did a lot of work on platforms a long time back — back in 1998, I invested in a company called WebOS that tried to go down the path of applying the desktop metaphor to the web, of duplicating the inadequacies of the desktop on the web. There were few people comparing last week's announcement to Adobe's Apollo — Apollo is setup to be a more traditional, extensible platform. One of the companies I am working with — im in like with you — is developing much of its service in Apollo. Apollo is truly a web application environment — offering state management outside of the browser, for example Apollo will let me do my web mail while I am unconnected. But Adobe is building this as a platform service, like Flash the intent is to proliferate the tool set across the web, developers will adopt it as will end users and like Flash it will provide revenue from scaled developers paying Adobe a license fee. This is a platform business model that the market understands. A cross platform run time isnt as sexy sounding at F8, but it might be more meaningful. And then there is Firefox 3 — another valid comparison that didnt seem to come up in many discussions.
3. How will application providers be promoted in Facebook? This is critical to understanding the underlying business terms between the distributor and the application creator. Last weeks announcement was about distribution, and it formalized an approach for Facebook partners, business development in a box, a highly scalable approach to partnering. But what are the underlying economic drivers? At AOL promotion and positioning was usually governed by dollars spent. At Google it now seems to be about long term strategic value: years ago the Google services that were tiled above search results – were best in class – for finance related searches (search for a stock ticker), Yahoo finance was promoted, Mapquest was the default when you searched for a location. Then slowly over time Google services received prominence equal or better to others. Today its pretty much all Google services upfront, in default positions — nice to leave some pointers for competitors but as Google knows well defaults drive traffic and traffic drives revenue.
Last week the COO at Facebook, Owen Van Natta, said: "How are we promising not to trump your application? We're going to level the playing field, developers won't be second-class citizens–we're going to compete directly with them." Accordingly, the Facebook application directory is organized today mostly by popularity — but mostly is different to always.
See the ringed sections of the screenshot — unlike third parties Facebook applications don't list the number of users of its applications (Marketplace is a Facebook application). And note the that Application directory (boxed) starts with Facebook's top Applications. Finally, as the users expands and contracts the application list (the more carat, where the arrow is pointing) Facebook's one advertisement on the page moves down, partially below the fold. Tell me this execution isn't setup to collide with business priorities.
In Japan, on the cell phone, Do Co Mo understood that with a limited UI placement of third party services needed to be ranked by usage. Is Facebook headed down the same path — and what does the COO really mean?? — Facebook owns this garden, competing directly with application providers is going to be, interesting.
5. Lastly, how does Zuckerberg social graph extend beyond the core college audience / behavior? The feed feature added a whole new dimension to Facebook and extended the time people were spending on the site significantly, Comscore data suggests it went up by over 5 mins per day. Fotolog has a similar, feature that alerts users to new uploads by friends — its a significant driver of our navigational based traffic. But how does the audience and the use cases evolve beyond the core? Will people outside of college enter in real names into profiles and will the social dynamics of the broader audience fit with the services that were built for the student based audience? Over the past year I have started to use LinkedIn more — its starting to become useful, the network is large enough, the alerts I get from LinkedIn are useful — not spam. I signed up for Facebook shortly after they opened up — but I didn't go back, till friends started inviting me. Over the past 6 months I have visited the sites to confirm friends but there is nothing useful about Facebook as yet, and useful aside it better be either personal or entertaining — but like so many other social networks its about collecting connections, but whats are the services that are going to drive usage for me — I don't see it yet.
This is a quote from Giga Om's review post the launch event, its worth a slow read. "Zuckerberg says you can serve ads on your app pages and keep all the revenue, sell them yourselves or use a network, and process transactions within the site, keeping all the revenue without diverting users off Facebook. This was the opposite to what was stated in the WSJ article earlier this week, and gets by far the biggest reaction from the crowd."
This got the biggest reaction from the crowd?? Maybe a crowd packed with Web 2.0 service and feature developers who are in need of an audience found it it interesting. If a user today opt's in to use your site on Firefox — or your application on windows — or even within the grandfather of walled garden's AOL — you still get to keep the ad-revenue. So why is this a big surprise? Maybe the attention the announcement garnered is also about the proliferation of web based features searching for a destination to marry themselves to.
Intent and that Telegraph Road
Walking thirty miles with a pack on his back
And he put down his load where he thought it was the best
Made a home in the wilderness
I do think its worth do ask whats the intent behind the Facebook announcement, who is meant to serve and whats the need behind the F8 initiative? The Facebook was launched as a service for US college students. It was full of social tools, it let you build out your own network, post events, notes, photos and most importantly its all private, so that students can develop a profile that is real vs. many of the fantasy based profiling you see on Myspace and other sites. Facebook achieved a lot of its early traction for the same reason as Cyworld did– you could enter your College, your year and actually find friends, colleagues, friends to be, cruches etc. Because people used real names on the service — emails were verified by domain and you could find anyone in your university. This was and is a big idea — few sites have a relationship based with their users that maps to real identities. Anyone who has attended a US university or college knows exactly what this is about. Then came the monetization.
Facebook started with advertising, they achieved some remarkable successes by mid 2005 they became profitable, they had 2,000+ colleges and 20,000+ high schools on the service. And the audience was rabidly engaged — 2/3rd's of the active membership came to the site everyday. But look at Facebook's reach through 2006 — it is flat, because by 2006 they had tapped into an audience and grown the business about as far as it could go given its natural limitations: students. They were now faced with the question of how to scale your business beyond its base. They could go global — there are services like FriendsReunited in the UK and Australia who are demonstrating, albeit with differences , that the market exists outside of the US for a Facebook like service. And /or they could opt to extend the scope of the Facebook offering and try to reach a broader audience in the US beyond students. They decided to push on both fronts but most significantly in September last year Facebook opened up to users irrespective of whether they were in school or not. In 2007 Facebook's reach more than tripled. Before they opened up the doors to the broader audience they were adding 15,000 members a day, today they are adding 100,000 a day (NYT stat, note Fortune says 150,000 a day). They now have 24M active users, posting mostly Photos, notes and events.
Then came the lawyers then came the rules
Then came the trains and the trucks with their loads
And the dirty old track was the telegraph road
But now reach has extended they need to find ways to get people to spend more time on the site. Here comes the platform initiative. The platform that was released last week is about extending Facebook in a different manner to the other social networking sites. Its about continuing to extend Facebook features by offering distribution of third party applications on Facebook. Yet the features been added are contained within the Facebook experience. Out the gate its a great opportunity for fledgling sites, particularly sites that are more of a feature than a destination — Facebook is offering one click installs for applications within Facebook. Its about distribution and its about continuing to drive the amount of time people are spending on the site, which in turns drives advertising. Facebook is playing the same game as media aggregators have played since the dawn of time. Whether its Disney, Yahoo or AOL — its all about getting in front of the distribution firehose — they are selling their audience. Day 1 its not setup as a sale. Remember that AOL used to pay service providers to offer content and services within the walled garden — then in 1996 when AOL hit a scale it stopped paying providers and started charging — bit by bit AOL flipped the model. This all seems far less interesting and ambitious than the headlines suggest. Zuckerberg told Kirkpatrick that what Facebook is unveiling would be "the most powerful distribution mechanism that's been created in a generation." I hope its is more than that. If Facebook's F8 is about trying to extend the size and scale of innovation and services in what amounts to another a walled garden experience it will another building block in the long history of web hype. The Facebook has a great social platform to build off, I hope they are brave enough to let their users take their data and extend services beyond their control, beyond the walled garden.
A last point worth making is the absence of Microsoft, Yahoo, Ebay and AOL in the platform / social networking space. Live.com was meant to be a web development platform — but things hewed back to Windows with the launch of Vista. Microsoft developed much of the thinking behind the web as a platform — with hailstorm and then live.com — but IE7 and Live haven't taken the lead. Yahoo made all these great acquisitions, many of which they they have left in silos and failed to build upon. Ebay has this amazing social / trust network that links merchants and end users. We think of profiles as been specific to social net, but Ebays profiles as they relate to trust and commerce and communications (skype) are a trove of data that could be opened up to users, applications and the web as a whole. And the merchant relationships, what about extending them into advertising. Like wise with AOL — there was a recent comment about the importance of opening up AIM, again… Its amazing to see the leaders of earlier generations of the web MIA — gone from this social networking race.
The semantic web needs to be distributed at its core, another walled garden is too low a bar for a really powerful and interesting social network to aim for. I hope Facebook actually step beyond the marketing hype and deliver a social platform for the web.