Nick ONeill (Social Times), Mike Sego ((fluff)Friends), Keith Schacht (42 Friends LLC), Tim O’Shaughnessy (Hungry Machine), Zach Allia (Free Gifts) spoke on a panel about user engagement at GSP West 2008.

How to measure engagement? At first, Facebook reported installs, now it’s number of average daily user. General concensus that this is somewhat misleading.

Are you getting lots of daily visits, lots of new users with churn or are you getting frequent return visits? Time spent on site, frequency of revisit, page views, content generation are also all good measures.

Why do some applications stick vs fizzle? Active work on the app, freshness and up to date features help. Things that provide an ongoing service tend to stick, one time use tends to fizzle. Freshness helps, people don’t want to revisit the same old stuff.

Little Green Patch held out as an application that has done a great job of always giving you something to do next. PacRat trading card game – amazing art, with really original gaming aspects that drive use. Someone else said PacRat is way too hard to learn. Bumpersticker is another application that has been great for users. It’s user driven content that drives freshness that scales.

Building gaming effects into the applications really helps drive engagement. Levels, leader boards. Well designed asynchronous game aspects can be really helpful in driving lots of use. Scrabulous does this really well.

Tim observed that if you can get people to use your application as a communication channel, you have reached the pinnacle of engagement. Reading Social was provided as an example of this.

User generated virtual goods that allow people to be raised up a celebrities within the community it can also drive deep loyalty and engagement.

New features that activated community – Mike said that gifting functionality allowed users to engage with each other in new ways. It created a user to user currency that allows the currency to thrive. Also, allow people to build up something over time. Leaderboards are another example of building things up over time. Give people the tools to add features, like quizes. Adding games into the application also enhances the user experience. Announcing today’s birthdays can drive birthday cake giving. Zach monitors this gift giving as his main metric of success.

Justin Smith (Product Manager, Watercooler), Andrew Chen (Futuristic Play), Blake Commagere (Mogad.com), David Gentzel (SocialMedia), Jia Shen (RockYou) spoke on a panel today about designing viral applications.

Andrew started the conversation by describing viral marketing as a marketing system where your customers sell your next generation of customers. Jia pointed out that the time frame now has been collapsed by the social networks so there is an accelerated viral opportunity.

There’s been a long evolution of viral from word of mouth, through email and other tools that have been turned into features of the social networks. Now instead of starting with the product, you can start from the customer and work back through the social networks as distribution channels.

Jai pointed out that early on Facebook did not put much restraint on how many invitations could be sent which created a gold rush effect that allowed an eco-system to grow as developers chased the growth. Now Facebook has put constraints on viral marketing tools which will make it much harder for new players to grow. The newer APIs are also being more conservative and that means they may have more difficulty building the same kind of ecosystem.

The differences in functionality across sites drives difference in application strategy. For example, on MySpace the focus will be more on applications that are self-expression, canvas oriented and less the viral, messaging applications.

Andrew highlighted that you can learn a lot from games in helping to make applications more successful. Things like reward schedules can drive use.

If you can build something that catches on you will know after the first couple of thousand users. If it’s successful with this group you can be confident that you can grow the applications penetration. Less viral applications will take marketing money to grow.

Being viral is not all it takes. If you’re getting lots of visitors and trials, and are not converting them into active users you are missing the opportunity. You need to make sure you’re measuring retention and repeat use of different visitor cohorts to ensure long term success.

It takes both viral growth and engagement to be successful.

[tags]gspwest08, graphing social, viral applications, viral marketing[/tags]

Roger Magoulas of O’Reilly presented some statistics they have collected regarding Facebook Applications.

Application count is up from 500 new applications per week to over 600 new applications per week, climbing up to over what appears to be 17,000 applications. The top 5-10 applications have really been the most successful in terms of usage, with a steep curve downwards for many other apps. 1% of apps account for 77% of volume, 2% account for 86% of volume, 10% account for 98% of volume and 20% account for 99% of volume.

Update: Found a Facebook Analytics site with some good date here.

[tags]gspwest08, Graphing Social, Facebook statistics[/tags]

Jeremy Liew (Lightspeed Venture Patners), Mark Pincus (Zynga Game Networks), Shervin Pishevar (Social Gaming Network), Michael Lazerow (Buddy Media), John Hwang (Tripmonger) spoke at GSP West about social games.

The panel talked about what they are doing and how they are making money. It all starts with how social games are different?

The opportunity is to create the very personal social gaming experience that we see when families play the Wii. Creating games like this online is the goal of social games. It’s about bringing people together, whether they be strangers or friends or family. The psychology of the games becomes more important in social games. The perspective is on an easy way for people to hang out together, not on game technology.

Experiencing the internet together is more fun than doing it on your own. A good rule of thumb is, if it can’t be made better using Facebook don’t do it. Users enjoy the light weight nature of these apps. Social gaming takes share from casual gaming.

Scarbulous is the breakout game. It’s so simple that it allows you to interact with lots of people in a light weight asynchronous interaction. Long playing synchronous games are played in a single session that doesn’t create a reason to come back. Asynchronous games give you a reason to come back. This can cause powerful viral effects.

Branded games are the key to monetization right now. It’s highly efficient way to get major brands and media companies reach their audiences. Average deals right now are about $100,000 for Buddy Media and they are sold out of inventory right now, so they are working on monetizing other company’s games. Another great opportunity for monetization is virtual goods. The amazing thing is that all the companies on the panel were profitable and they think there is still more upside if they can close brand deals.

Are social games a hit driven business? Games are hit driven. The big ones get bigger and the rest fall off. The is definitely a cycle of growth, use and then fatigue. The end level though may be higher than what is seen elsewhere on the web. A number on the panel are working on how to create networks to help leverage existing installed base of applications, either through click networks, APIs, partnerships, platforms that can be rebranded or acquisitions.

[tags]gspwest08, graphing social, social games[/tags]

Jim Benedetto, Vice President of Technology for MySpace, presented a technical overview of MySpace’s Developer Platform. The presentation covered where MySpace is now and where they are headed in opening up the MySpace API.

MySpace has had a history of a strong ecosystem

• Youtube , Photobucket, Slide, RockYou are all examples

MySpace values are driving there developer platform.

• Minimal creative restrictions
• Encouraging self expression
• High level of customizability
• Beneficial to user, developers

MySpace API overview

OpenSocial APIs

• javascript/html for embedded apps
• with myspace extensions

REST APIs

• server to server comm
• Oath authentication

Actionscript APIs

• flash support

Why OpenSocial?

Openness

• openness key to Myspace success
• openess helps everyone

Portability

• developers can spend more time building a great product rather than rebuilding for every social network

Leveraged existing technologies

• no need to lerrn proprietary development languages

OpenSocial Support

• full support of public spec
• currently support V0.6 with support for V0.7 soon
• MySpace specific extensions for MySpace specific features
  • bulletins
  • additional attributes for bands

Platform surface

• Profile surface
• Canvas surface
• User homepage surface (private space)
  • User specific surface
  • Enables the application to show specific data to a user
  • eg ebay bids, tweets from friends
• Application Gallery
• Application Profile

Security, Privacy and Safety
Applications will go through safety review process makes sure it doen’t spread bad or rogue code or viruses, keeps data secure and protects against rogue actions from other users

Apps will be governed by same privacy controls that are in place for members
will be using new technology to ensure that applications are safe for end ocnsumers

• Caja – MySpace and Google Joint Javascript Sanitizer
• Proprietary MySpace Technology

Balancing Virality and user Experience

Virality vs User Experience

• take long term approach to growth and distribution
• ensure clean applicaiton experience

Spam based growth is not needed for good applications

Measured approach to application growth

• Workflow limited to sending message 1 to 1

Restrictive early on, slowly increase communications channel

[tags]gspwest08, graphing social, MySpace[/tags]

Patrick Chanezon (Google), Chris Schalk (Google), Kevin Marks (Google), Lou Moore (hi5 Networks) spoke on a panel about building OpenSocial applications at Graphing Social Patterns West.

What does Social mean? We look at each other, talking, laughing, we help each other, we read together, we do projects together. We also have social objects that we tell stories about.

The challenge is how do we socialize objects online without having to create yet another social network?

OpenSocial designed to do this. OpenSocial is one API that works with many websites. Just now getting through gartner technology hype cycle over the last 3 months. OpenSocial is now on version 0.7

Core OpenSocial services include people, activities and persistence. After a tour through some of the code used to make the API work, Kevin Marks described two open source projects designed to help build appliations – Caja is project designed to help prevent things from going bad. Caja is an optional Javascript sanitizer for use by developers. The panel recommended using Caja or at least tesing using Caja.

The second project is Shindig. Shindig is a reference implementation of the OpenSocial and Gadgets stack that is designed to accelerate development and deployment. The panel said that now is the time to start developing.

Update: Here is the slide show:

| View | Upload your own

And a post by panel member.

[tags]gspwest08, graphing social, opensocial[/tags]

David Glazer, Director of Engineering at Google and leader of Google’s OpenSocial team presented at Graphing Social. He started with a perspective that the cloud is here.

Fast, easy, open, and everywhere are key reasons why cloud computing wins. The cloud is about getting the computer out of the way so that we can be more productive.

How does this matter in the social cloud? Getting the computer out of the way so that we can interact more easily and be more productive. People are the killer application of the web. This is not new. Getting me to the information I want has a long history. email, newspapers, bookstores, ftp, gopher, bbs all were the beginning of this. Social used to be spelled c-o-l-l-a-b-o-r-a-t-i-o-n.

Need to get the accidental barriers out of the way.

When Google thought about social they started by asking where would it be better to do things with friends. ie add “do it with friends” to end of sentence such as – You will plan a trip with your friends.

Hard problems still out there. These are accidentalt barriers.

Fragmented authentication is a problem. Different logins across sites cause people problems. OpenID is a good start, but has a ways to go.

Fragmented social connections. Too many places to set up friend.

Fragmented applications across sites – not the same everywhere my friends are.

Goal is tomake the social web better. The end result any app, any site, andy friends without having to think about it.

First step is to share the pain.

OpenSocial’s goal is to allow you to

• Invent it
  • xxx with your friends
• build it
  • standard web app (html/javascript)
  • new JS APIs(who I am, who I know, what I do)
• Run it
  • on any social site that runs api

There were a number of surprises from launching OpenSocial:

1. Breadth of interest. Classic social networks wanted to use it, the business oriented sites wanted to use it, enterprise software wanted to use it, communities wanted to use it. Lots of different types of uses, all with the same basic use case.
2. Open source is a good idea. Shindig in Apache incubator was set up, thanks to Brian McCallister at Ning. Clear mission, open license, engaged community and real world use.
3. Execution is hard. The idea is ahead of the execution. This is what’s next.
4. Questions about can you also help with… connections across networks that extend the solution to other aspects of connecting people, apps and sites.

What’s next?

1. Get the spec and build apps
2. Contribute to Shindig
3. Join the group and grow the spec

[tags]gspwest08, graphing social, OpenSocial[/tags]

O’reilly’s Graphing Social Patterns West they had a speed demo night where developers get to demo apps. They ran through 10 applications. Many of them are from Facebook. We SMS voted for the best of the first six and then again for the second four demos.

Group One – Facebook Apps

• Who Has The Biggest Brain? Facebook App, 600k downloads, 4 minutes per session – 2 million minutes 22%
• Just Three Words 7 minutes on site with lots 2 hours per day three weeks to write two weeks to rewrite 3 million words, 60k stories, average size 200 words, created for public – 2%
• Puzzle Messages – simple hybrid between puzzle and messenger have total of 27 apps and 12 million users – 1%
• Ski and Snowboarding – 1200 resorts in the world – Where have you been? Show off where you have been, share photos etc. Users have geocoded where the resorts are, correcting a lot of what had been poor data about location. -9%
• Dipity – Wikipedia for timelines in Facebook – takes all the data and connects other people and publishes the results in a timeline. Allows you to add other RSS feeds to timeline. -22%
• Developer Analytics – Analytics for Facebook Apps. There’s an analytic tab that’s hidden right now. Virality(install as a result of some one else installing), Engagement(return uses), Revenue. Rolling out to private beta in next few weeks. -40%

Group 2

• Know your neighbor – Written for Orkut – Twitter for your community and neighbors on steroids. Sorts neighbors by distance from you and lets you send messages to your neighbors – 6%
• Reading Social – built on OpenSocial sharing about books across platforms. Also showed dining social, drinking social, music social – 41%
• Going Places – connecting people and places – social travel application. “Smart Travel Layer” connects to Facebook, MySpace, Orkut. Take test – match with friends on compatibility basis. Add places you’ve been. 400,000 places in database. You can see who has been to places you want to go and see how they rated that place. – 8%
• ChirpScreen – photo sharing app using a screen saver. Facebook, Twitter and Flickr accounts all feed the app. Has comments back to photo owner. – 43%

Note: final votes will be updated tomorrow.

[tags]gspwest08, graphing social, GSP AppNite[/tags]

Chris Messina presented about DiSo.

DiSo is a project founded by Steve Ivy and Chris Messina, built on top of WordPress. The project aims to explore the design of a distributed social network using many of the building blocks the blog software already supports, while leveraging technologies like XMPP and XFN for friendslist federation and message delivery.

Social networking is not all about the add. It’s not about stupid apps. It started out with niches, MySpace – bands, Facebook – college students, LinkedIn – professionals. They were all system centric views of social networks. This causes the view of system centric value, not people centric value. The internet is much bigger than any sytem out there. We should not consider users to be customers.

We need to move from system centric view to a citizen centric view of the world. Who are web citizens? They have identities, they have provenance, they have friends, they have agency.

What are the aspects of a distributed social network. It has a home for your data. It has the ability to manage the data. Systems should subscribe to me. I should not have to subscribe to them. There needs to be flexible permission systems.

Benefits from a distributed model. Cross pollination of value from different systems. A more up to date profile that makes it more valuable to other systems. Permission for everyone to build based upon the profile. Easier upgrade path.

Components of a DiSo site. Activity feeds are proliferating. Ability to publish friends list and ability to move it around. Messaging and notifications are critical components. Filters to handle flood of messages. Standardization of terms and permissions. Groups, grouping and events that put the user in charge. Identity consolidation and rel-me.

Technologies that are available: OPenID, Microformats, OAuth, XRDS-Simple,ATOM/APP, Jabber/XMPP

Current status: Starting with WordPress because it’s PHP and there was an existing OpenID plugin they could use. You can find out more at Diso-project.org

Update: Here are the slides

| View | Upload your own

[tags]gspwest08, graphing social, DiSo, Chris Messina[/tags]

Social networks are letting more people connect than ever before. With just a click, you can make friends with people around the world, and share your work, hobbies, and other interests. But how visible is that information?

Dan Farber (CNET Networks), Joseph Smarr (Plaxo), David Lavenda (WorkLight), Allen Hurff (MySpace), Ben Metcalfe (Swordfish Corp / DataPortability.org) discuss managing privacy and data portability for social networks.

Where are we on data portability?
We are still in a walled garden environment. I may be a Digg user, but finding all the other people I know is a pain. People keep scraping your contact list to find other people. We need to get both the technical and the human side of this to work.

Discussions about online reputation, who owns comments etc need to be taken into consideration with respect to how you support data portability and still manage privacy.

What do we need to do to bring intranet data to social sites?
On the data side is the primary issue is how to manage security and access to data. On the work flow side, control over data and data security is a big issue. Secure Enterprise 2.0 initiative is just now getting up and running to focus on the issue.

Where are we on the Data Portability Initiative?
Companies recognize it is a big problem. It has taken time to pull the group and organization together. A number of start-ups will announce when they release products. Across the board there is support, but it is taking time. Are companies still protecting there data or will those that go open gain an advantage? Still an open question, but these large companies want to be involved to help shape how this concept will scale.

How do we get consumers to grok this?
Build real use cases that the consumers understand. When you show them in a way that makes sense to them. Who gets to see your photos etc? Once consumers see it working the right way somewhere they will point to it as an example of what they like.

[tags]gspwest08, graphing social, data portability, privacy[/tags]