Saturday, August 25, 2007

Facebook Developers Garage - Build on FB as Platform

Facebook Developers Garage on Saturday Aug 25th 2007. rocked!!!! It was an awesome event with 300 developers all handon discussions at Facebook headquarters, with excitement in the air that we are all part of some new beginning.

Caitlin O'Farell of Facebook organized it and keep a smooth flow of the event. Thanks Caitlin!

10 min of Lightning Round presentations by real facebook App developers
Everyone sharing their experience building and scaling Apps was the highlight of the event as they had different viewpoints about scaling, my summary is below.

All in all, Facebook is in early stages as a development platform with great promise with many smart minds betting on it with lot of promise to bring the real promise of the web to provided hosted solutions with power to user communities and promise of building inter-operable applications building on each other, reducing development costs. Its left to us to build Applications that create fundamental value and create businesses here!

1. Tyler Ballance of Slide was a great speaker, spoke in depth about building apps thats scale:

* Choose the language you are most comfortable in - php, .net or rails Do not choose php because Facebook is built in php or rails because thats new.
* Focus on building App and getting it to users, don't write spagetti code, optimize later.
* Compartmentalize your business logic from interface layer, similar to any large software app during design.
* Facebook helps to scale Apps using social graph, so you may scale from 1 to 3Mil in a week, so focus on your database which can crash with too many hits, keep a simple schema, key on integers,
* FBML on canvas pages is easy to scale, it puts the load on Facebook servers and they fix quicker than App developers, your App doesn't have to fetch users name, profile etc. Cut unneccssay calls to Facebook API because if anything on facebook users blame you as they cannot tell the difference. eg. iframe
* With IFRAME/HTTP you never have absolute control. The disadvantage of using IFRAME is the need to call FB API a lot and it heightens load on your server. Also limitation on multi user access with FBfriend selecter so, Slide (tyler) built their own multi-friend seletor for Top Friends App.
* Use Browers processing to your advantage eg Flash on the browser, use Javascript even if slow, offers better user experience
* Keep in touch with new FB stuff, they are fixing and adding to our feedback.
* Most users are on FB for fun, checkout what friends are doing, leisure activity. integrate as much as possible with social graph.
* Someone asked about importing 3000 friends. Tyler said there is technical limitations in interacting with FB API they'll timeout so they did batch loads in Top friends.

2. Joyce of Booze-mail from Renkoo $1M users serving 1Million drinks served per day.

* Joyce had a different views from Type on software calls and perception of Facebook users behavior.
* She gave detailed of what backend they have to manage the load.
they use HAProxy as Loadbalancer- free open source one and 3 webservers on Amazon Web Services which is slow, old, not too much memory with problem for persistence.
* They use Heap Writes, denormalizing them.
* Added a tier of Image caching, They use "Vanish" not "MemcacheT" which was Tyler's favorite (so it appeared)
* Joyce thinks people will do more than the simple fun things of today on facebook, though Booze-mail is purely in that fun category.

3. Matt Culver of Amazon Web Services(AWS) spoke briefly about using AWS, it first appeared he was part of the Booze-mail team, but he was talking about Amazon Web Services for FB App developers.
* Most people in the room knew about Amazon Web Services so he just gave some pointers.
*Amazon EC2 - blog, wiki - and CTOS Who Use AWS, a group on Facebook. Interestingly this group has open discussions on AWS alternative and comparision with AWS.

4. Ali Partovi, CEO of affirmed how Facebook was THE PLATFORM for anyone building any web business.
*They started as a dot come with and their growth spiked once they decided to become a total facebook application business.
* They have 16 Engg, 100% of time spent on Facebook Platform
*Ali is very optimistic on Facebook as a platform and addressed some user's concerns. He said think of creating an entire business, not just another widget.
*Are you worried about dependency on Facebook?
He trusts current management team but says build interdependency with FB depends on us too so they won't compete.
* He said all business on any platform needs to think of sustainability so no one can replicate their business. Ilikes is building connection between parties - itunes and facebook, ticketmaster, platform for artists to create profiles on facebook. They offer a music api to add music to other apps which is stitching them into the Facebook ecosystem.

5. Courses - Michael and Joeseph spoke very honestly about their experience.
*Wish I could talk to myself the night I launched.
*Take lot of advice from slide.
*Common theme of the day from everyone was- put your app out there and see what works.
*Get mathametical with your app - track every link - where its coming from, every invitations, see whats works, whats not, adapt quickly to emphasise the stuff that works.
* Settings preview of profile boxes - let users know as users get desensitized by notification.
*Michael was teacher for 3 yrs, found it difficult to get students to get to use any technology, but students were connecting on Facebook so this was perfect to build course on FB.
*Users care about things you didn't know so listen to your discussion boards.
* He said "Slide is wrong, not just fun, facebook is going to become the center of how we organize different aspects of our life."
Personally it reminds me of the early days of the web when everyone defined it from their perspective.

*What rail server - mongrols, optimizing for scaling have lot of work to do! have 30K users. *Working on a tax application by the side.
* Mistakes -didn't launch quickly once Fb decided to get rid of their courses app.
Focus on one thing that your users care about.
*Two of us full-time older brother helping part-time at night.
*Fun - making invite page lot better. - video chatroom where people can get together and talk.
*Tracking - started by using Google Analytics, slowing us down, build your own analytics into app right at start.

6. Joe an Eric Diep - Quizzes interviews - 4 mil users

*Started at F8 launch party.
*Work independently - manage your time, set your own schedule;
*Wear a lot of hats. scale, develop, dedicate time to talk to people outside us,
*Don't worry about scale.
*Great opportunity to make something in 1 weeks effect.
*Your app is prototype, pay attention to user feedback, some user needs guidance.
*Track everything what works what doesn't.
* We have 4 boxes for servers
*How do u deal with imitiation app - with same name. Facebook allows app with same name which is an issue for all developers.
*important not to think about monetization -till u have users invested in your product, coming back.

7. James of

*James platform Director for
*Facebook - not going to be same in 2 years, getting ready for it.
*Users - learning - schools used CMS. connect learning in school and your social graph. *Edutainment games, tutoring, - all apps
*Opening Files App as a platform for developer on facebook.
*Apps to share docs and tags and description with other apps.
* Tyler compared with FB Datastore which came from Facebook. Looks like something Facebook launched similar to the App, which is going to be an issue if we build something that can be part of the FB core OS.

8. Rock You Lance Tokuda, CEO RockYou

* Gave good stats on distrinution of users for apps. 87% Apps have than 10K users
1% have Million+
*RockYou owns Super wall which has 4.8Mil users
* He gave a casestudy of how they helped Y! Music Videos scale.
Y! Music App had only 7K users in 1 month. RockYou ported same source code and added viral features got 900K users in couple weeks.
What they did:
- Cross sell apps network got 100K users
- Virality , integrated with superwall,
* Lance gave a framework of designing good Facebook Apps
a) New users - simple linear flow, min steps to configure apps, provide defaults (min work)
b)Engage friends
Engage on invite- strong call for action
Ideal is when friends want to engage all friends
ongoing engagement - trigger events to invite friends or connect with them.
c) virality
Strong call to action for receiver
d)Integration with other Apps - cross sell existing user base
eg. likeness - quiz related to music

peter louis who developed superwall was there to talk to everyone.

* Louis announced Superwall API - All media types supported images, video, flash.
9. Chris Chan - Causes 3M users.
*Decision - name causes has a positive feel.
*Looks like Facebook (not separate branding), credibility with same font size, style.
* Decided to accept donations
*Viral flow - show cause page without adding app. If you shorten viral play too much, users won't understand what the apps is about.
* Tech - Asynchronous, reduce depenedency on Facebook. Stats we track and update, separate from user requests. Hired SQL consultant helped.
*F8 Story - Blake built zombie and vampires and chris. 10 days launched at F8 at 2pm.
$300K for donations, its 10cents per user. Testing is like eating vegetables.

User Announcements

1. VC view (didn't catch the VCs name, will check my notes and update): People will come, Facebook users are going to be worth $10/user so everyone should build apps for facebook.

2. Worth checking out

Dave @ Facebook summarized the day, then there was Pizza, networking and a round of indepth technical discussions.

Dave's summary talks about Facebook's welcoming attitude towards users and builds trust to build on Facebook as the next platform.

He says, users drive growth for apps, listening to community and adapt. Some apps are building an ecosystem of services offering interoperability between communities, which is very exciting for Facebook, which he calls "Value center community" so he encouraged everyone to build a suite of apps like Rockyou.

PS: I'll upload my photos from the event on FB later.

Friday, August 24, 2007

Animoto - Making videos in 5 minutes Awesome

Found this cool technology to make videos so easily in 5 minutes!

A new startup Animoto, based out of NY has come with a cool technology.

Its child play - What is video - you pick some pictures, add some music mix them, make the pictures move and become a video! Add 15 pictures and in 5 minutes you get a video!

Awesome!!! Personally I was prepared for funky UI and no help and Animoto pleasantly surprised me. Its clean UI, seamless experience with decent help message on the side.

I emailed them and got good support response too.

You can see the depth of thinking behind the technology and productizing it. They offer the choice to use your own images, or pick it up from Flickr or anywhere on the web. They allow you to post the video on any blog or myspace etc. I am going to add this to my Facebook for sure!

Only limitation now is that they seamlessly log into my Flickr account but don't allow me to pickup public Flickr pictures to make a video. Appears more a decision not to get to get into copyright issues!

They have free 30 sec video (you can get a feel for BarCamp with this video), or you can upgrade to make longer movies. They allow private sharing of videos or uploading for public sharing. Intutive interface instead of the standard public and private views! They have teasers of what all is coming in the pipeline so you can take the video to your ipod or burn a CD I guess.

Animoto has great promise, has lot of startup energy, smells of success and fun. Sure viral play, so we are going to see a lot of Animotos soon. Enjoy!

Saturday, August 18, 2007

BarCampBlock - Thank You! Tara you are Ms.BarCamp!

I've been to BarCamp Boston and Manchester, but it was a totally new experience going to BarCampBlock in Palo Alto, where it all started.

I spent more time on the block than sessions and learned a lot! The beauty of BarCamp is that each person takes something different based on what they are focused on and what they contribute!

Great Organization to the team and Ms.BarCamp Tara!
My previous topics at Barcamp Boston has been about taking a concept to business and raising money. I decided to go with an open mind and not talk about the same topic.

Thanks to Tara, Chris, Tantek, Ross and Liz on great organization. The energy all over the block was amazing and contagious. Tara would be talking to a group, and keep an eye to refill drinking water and food to keep us all fed. She is amazing, and is Ms.BarCamp! captures the chaos and energy as everyone rushed to the grid to build the conference schedule on the fly!

Brian Solis has done justice to cover the people of the crowd! Missed Robert Scoble sneaking out of his non-blog week :-)

Lawrence Sinclair was a star as he instantly brought together a bunch of Facebook App Developers and formed a co-working group to develop more cool Apps.

Its an understatement to say that I met most of the people who I've got to know in the first 6 months of my valley visits all here at BarCampBlock.

One has to experience BarCampBlock to know how being a techi is cool here, being an entrepreneur is great, and everyone gets to action to brainstorm and collaborate in real time to create the thriving silicon valley culture!

I've been bitten by the "Facebook As a Platform" bug recently and some of Dave McClure's tests on Facebook and it influenced my view of what hallway sessions and BarCamp sessions I participated in.

Sessions - So much to choose, so little time

So many sessions covering so many topics! Too bad I don't have a Time turner like Hermoine so, can't time travel to attend two parallel sessions.

I presented a session on "Social Media Optimization (SMO) for Facebook" which had 40+ audience in the Socialtext co-working room, with a whiteboard and great conversations.

The Wikipedia definition of SMO is broad. I focused on SMO for Facebook App Developers. What to utilize of current Facebook API and support to make your App on Facebook to reach users with clean UI, messaging, target users and influencers of your space and create viral play.

I had fun and left with more things to think about and made more friends on Facebook :-)

Session Summaries

Nelz has done a great job of summarizing sessions I missed!

I liked the session on "Future of Widgets" by Shawn Van Itersum, Product Manager of Clearspring, a Widget development platform.

He had a session on intro to Widget in the morning. The afternoon one had great conversations. Rashmi of had great questions on widget analytics which are issues I have faced.

Clearspring offers a Widget development platform for developers. I am not sure how it compares with going with RockYou for Widgets or the Gigya Widget distribution platform. Shawn was a good presenter and made a compelling case for Clearspring so I am going to check them out for sure!

The discussion was not about his company but about widgets and problems of supporting widgets for multiple platforms, distribution etc. Clearspring is mostly available for content widgets on Myspace. There was some discussion about Widgets for the blogosphere.

All in all, a day spent well, worth pondering year long till next years's BarCamp.

Update: Brad Templeton of EEF was there. Gotto join EEF to donate using causes App.
Had great conversations with Nate (CTO) of Pbwiki.

Friday, August 17, 2007

BarCamp Block Sat and Sun

I am going to BarCamp Block in Palo Alto tommorrow (Sat), Tara Hunt seems to be doing all the work like at Web2Open.

I stepped in the world of BarCamps first at Boston BarCamp and helped organize one, spoke on technology commercialization on the fly, and went to Manchester BarCamp where I presented Building the Business Side of a Business.

This is the first one in the valley, Tara has said 900 have signedup, don't see much people on the wiki as they the attendees profiles are on Crowdvine, which I've not learned to use much yet but exited at the randomness and real time makeup of the conference!!!

Do say hello if you are at BarCamp tomorrow.

Facebook Seeded Entreprenerialism

I met 10 people at Facebook Developers Meetup last night at Neto in Palo Alto and saw 3 cool Apps not formal demos, but entreprenurial feedback sessions on future plans - Zoo by Action, ulinkx by Saket and a stealth mode professional networking App by two guys who moved from NY to launch a cool FB App . Dave Westwood had great experience from Myspace and helped us make the connection to Facebook as the new platform.

I came back excited remembering the day I first saw the Mozilla Browser and the Web revolution began.

For the past few months, I've had my share of Facebook hype stories and ideas for Facebook Apps for startups I am associated with and exposure to old business types analyzing like many analyzed the Web as it shifted and morphed and grew and crashed and was reborn.

Heres my new revelation why I believe Facebook with it semi-open platform is a real business here to stay and threaten the existing Web ecosystem only to help new entrepreneurs:

1. Facebook offers a UI platform with built-in viral play so anyone can build an App on FB and instantly get viral play within its network. The web is full of smart techies who want to build technology cooler than the next person, but no technology is sustainable business by itself.
So every App builder needs to market it. FB offers instant viral play just by building on its platform.

2. I think today, times are analogous to the early days of creating a web site when Yahoo was manually indexing them in their home page. My favorite FB App Booze-mail got 200K users within 2 weeks with no marketing.

3. Everyone loves Google for its cool search, but what made them a business and the biggest brand we all wake up to is their Ad business helping big and small businesses by bringing their customers to their sites. Facebook is a real threat to Google in that App developers are flocking to FB to park their new tech ideas to allow FB to send users their way to start building dependency on FB as we did with Google.

4. Theres an entrepreneur in every creative person, especially ones who can code as a way of expressing themselves. Facebook makes it cheap to build an App and get started with user traction. Not a full business, still a great start!

5. I have smart friends who are watching on the sidelines questioning the proprietary platform on limitations of access control and ability to personalize a user's experience in an Application environment. I agree the FB Apps I have seen so far (not including the ones I saw last night at the FB Developers Meetup and the App Cause) are playful ones as a test App built by smart techies.

Its the beginning, some Apps make an hybrid using the FB platform as the UI and lead to a Web site. Others are building robust backend systems with FB as the UI platform to benefit from the viral play.

Facebook has to open up a lot and make many important decisions to become a Google and not Netscape, but it looks promising that we can all be part of building the social operating system stitching our Apps to create a new ecosystem here.

Its exciting as the beginning of some real outlet for technical expertise with a business window to reach users to validate your idea. Welcome to Facebook Seeded Entrepreneurialism.

Be sure to checkout Dave McClure's comments on the hype around VCs investing in Facebook and the real venture fund focused on Facebook Apps - AppFactory run by Salil Deshpande.

Wednesday, August 15, 2007

DEMO Alumni and the Pleasure of a Startup Launch

I didn't realize I was part of yet another network of people connected by the tag "DEMO Alumni" till recently as there was a DEMO Alumni party last night:-)

I had launched Coola at DEMO in 2000. Launching at DEMO has been a life changing, career changing for sure, experience for me.

Chris Shipley has been working for 20 years identifying new technologies and providing a beautiful launch pad with DEMO conferences.

I have the greatest respect for her eye for spotting promising technologies and winning teams. You can see the list of successful companies who launched at DEMO and exited or IPOed. For those who do not exit their first startup, you can still see future success of the same founders.

Chris Shipley helps shape the business concept by asking smart questions, listening and helping the teams figure out their blind spots and package the company with a clear launch checklist which makes DEMO a fun and unique event and going to DEMO an exciting time for all DEMO companies.

Recently I've been involved in the DEMO process (hush, can't tell you anymore now) and am impressed at how well-oiled an engine it is with a weekly calendar listing deliverables to get ready to launch for DEMO. This will help the new kids from school who come with a cool technology to define launch with such a high bar set getting ready their product, web site, brand identity, press kits, getting ready with all aspects of launch marketing, DEMO presentation honing your pitch and Exhibit Stall to showcase your product and a media wish list for targeted PR.

I think of getting to present in DEMO as setting high your standards on the business side of your business and a good start for anybody's career.

My favorite DEMO Alumni serial entrepreneurs are Vincent Paquet and Craig Walker, who build Dialpad, sold to Y!, most recently build Grandcentral sold to Google. I've been a Grandcentral fan right from the start and noticed how they set about meticulously building their company, stitching themselves into the ecosystem, scaling site content to show company growth and incrementally improving user communications all DEMO style.

Hope to see you at DEMO again and next year at the DEMO party!