One of the biggest problems on the Web is joining a new social networking site. To do so means going through the painful effort of creating a profile and adding all of our friends, something we’ve done over and over…at least once for each social networking site we already belong to. This is quickly becoming an issue for everyone who uses social networks.

This problem has led to a flurry of activity, highlighted by LiveJournal creator Brad Fitzpatrick’s missive: Thoughts on the Social Graph, in which he clearly outlines the issues involved as well as some worthy goals to shoot for. Brad’s piece was followed shortly after by the Bill of Rights for Users of the Social Web, which among its rights is the right to allow users to syndicate their own profile and friend data. This, of course, would alleviate the squeaky wheel.

In addition there are countless groups getting together to try and solve this problem. The microformats folks are working on building formats to help with this. The DataSharingSummit is an entire event focused on this and related problems. All of this activity is centered around one idea: that people have a social graph that can be represented in software. In other words, we can recreate our offline relationships online and let everyone know about it by sharing some sort of file or feed.

The major axis of the social graph, as Fitzpatrick points out, is relationships between people, or more simply, a list of friends. My social graph, for example, consists of my friends, colleagues, family, and acquaintances. These people I know to some extent or another, some I talk with daily, some I know only online, and some I would rather not speak to.

At this point we could easily move forward and accept the common notion of social networks: that they are made up only of relationships between people. But for those folks working deeply on these issues another problem soon arises: the realization that there is more to the social graph than just people…there are objects that mediate our relationships as well.

I’ve always been uncomfortable with the word “mediate” because it is not entirely clear what it means. But in this case it means something like this: our relationships with other people are determined in part by the activities and objects we share. This idea has long been known in the world of social psychology, and could have big effects on the utility of the social graph going forward.

For example, our YouTube and MySpace and Flickr friends exist partly in relation to the content that we’ve shared with each other on those sites. Our lifelong friends exist in relation to the things we’ve done together: the places we’ve gone to, the words we’ve spoken, and the movies we’ve seen. It doesn’t make sense to talk about our friends without these mediating objects, and that’s why our social graph must also represent them as well.

This view is explained wonderfully by Jyri Engeström in this post: Why some social network services work and others don’t — Or: the case for object-centered sociality . Engeström argues for an “objects of sociality” view of social networks, where people aren’t the only objects necessary for relationships. Engeström’s post is in turn based on the work of sociology professor Karin Knorr Cetina.

This notion of “objects of sociality” helps explain the success of sites such as YouTube, Flickr, and Netflix. (and, I might add, the slideshow sharing service Slideshare co-created by my friend Rashmi Sinha, whom I first heard the term “objects of sociality” from). What these services have done is to create a system that supports relationships around the objects of videos, photos, and movies, and slideshows. And as I wrote about the other day (What if YouTube was simply lucky?), their success seems based on their ability to make the activities of uploading, viewing, and sharing as painless as possible.

Still, Jyri suggests that most notions of social networks are restricted to people. He says:

“Approaching sociality as object-centered is to suggest that when it becomes easy to create digital instances of the object, the online services for networking on, through, and around that object will emerge too. Social network theory fails to recognise such real-world dynamics because its notion of sociality is limited to just people.”

It’s interesting to note that Facebook, who is at the center of this social graph discussion because of their partially-closed (or partially-open, however you want to look at it) system, asks everyone who adds a friend how you know them. They want to know if you worked with them, if you went to school with them, or if you met them through an acquaintance. These items, the job, the school, and the other friend, are the very objects of sociality that make the relationship work.

This is why: the insane growth of YouTube had a definite starting point…the release of Lazy Sunday. I knew that Lazy Sunday was a factor in their growth, but I didn’t realize how big a deal it was until I graphed it out on Alexa. (not that Alexa is the end-word, by any means, but even if it is somewhat accurate the graph it would still show Lazy Sunday as the starting point).

What if it was Simply Luck?

What if the viral growth of YouTube was luck? What if, for example, someone had uploaded Lazy Sunday to some other video service? Would that service have taken off and become #1 instead of YouTube? Was YouTube just the product of serendipity?

If the answer is yes, then that suggests there is another social component to viral growth of web services. A social component that goes beyond just the inherent capability of the system: beyond the features. It may have something to do with ease-of-use…the person who uploaded Lazy Sunday chose YouTube because it was easy. Or maybe they chose YouTube because they could share the video on their blog. It could have been the design of the site at its most basic.

Or, it could have been some other factor. Maybe the person had a cousin who worked for YouTube. Maybe they liked the color scheme. Who knows? We’ll probably never know…so where does that leave us?

I think it leaves us with the notion that social sites like YouTube are ecosystems…places where growth is possible but not guaranteed, and success usually depends on care and feeding over time vs. creating a fully-formed success. The best we can do is provide an environment where growth and sharing can occur with the best weather possible.

3 Primary Video Activities that YouTube Got Right

Here are three ways in which YouTube created an environment for growth:

1. Uploading: Everything to Flash video.
   Plays everywhere. No platform issues here. YouTube sucks up whatever format your video is in and transcodes it to Flash.
2. Watching: Videos play by themselves.
   The easiest activity is the one most done. Ever notice that YouTube videos play without you doing anything? Talk about making the primary purpose of the site easy…it just happens. I love this.
3. Sharing: Super sharing features.
   YouTube has always had great sharing features. They include the code to share on your blog right alongside the video. Again, you don’t have to do anything for the code to be there. You just copy it, and it just works.

I don’t want to play armchair quarterback and say “YouTube’s design was what led to their dramatic growth”. Obviously, the release of Lazy Sunday had a lot to do with it, so it wasn’t just their design or their marketing or their investors. None of those single factors made YouTube the success that it is.

But I do want to know why someone might choose YouTube over its competitors early on in the game, before network effects started to take hold. And, given the way that YouTube has really made the three primary video activities easy, I am willing to say that YouTube created a fertile ecosystem in which growth could happen…by tearing down the boundaries to uploading, watching, and sharing videos. And that’s probably the most we can ask of any social site these days. Create an ecosystem in which things can grow, and with a little (or a lot) luck it might just grow like YouTube.

It involves Lazy Sunday, the hilarious Saturday Night Live skit performed by Andy Samberg & Chris Parnell which exploded on the Web in January 2006, generating over 5 million views and generally signaled the power of the viral growth of video. (read more about Lazy Sunday here)

The stat: in the weeks following Lazy Sunday, YouTube’s traffic grew 83%. ¹

At the time YouTube was growing, but Lazy Sunday was, in today’s vernacular, the “tipping point” which shot YouTube to stratospheric mind share. After Lazy Sunday, there was no question as to who the #1 video site in the world was. (YouTube was purchased in November 2006 for 1.65 billion)

This tipping point is even more startling when you look at the growth curve of YouTube…and notice that it started really growing in the December/January 2006 time frame. Wow.

What is Viral Growth?

The story here is viral growth. We all have a sense as to what viral growth is, but do we really know how to make it happen? Let’s imagine that we were starting another YouTube-like service…how would we go about doing it?

In the case of YouTube and Lazy Sunday, there seems to be two pieces to the puzzle.

The first piece is great content. The quality of Lazy Sunday is what drove the viral growth of YouTube at that time. It’s just a great, funny, unique video that you can’t find anywhere else. It’s got hooks into several different audiences, as it references various pop-culture phenomena (Chronicles of Narnia, Google Maps, The Notebook, Ghost, getting high). This type of content drives real demand…people like it and want to talk about it.

But people also needed an easy way to share it. Interestingly, in addition to YouTube Lazy Sunday was also put on iTunes at the time. iTunes, of course, is the world’s leading online music distribution service. But the growth and spread of Lazy Sunday didn’t happen through iTunes, it happened through YouTube. One obvious hurdle is that iTunes charges for content. Another crucial difference is that iTunes sharing features are to YouTube like a Gremlin is to a Tesla.

There is where design comes in. YouTube has some great sharing features. They have fantastic Flash player that works everywhere as well as super-easy cut-and-paste code that bloggers can easily share with. It’s practically a no-brainer to share video with YouTube. This brings to mind the #1 rule of interface design: the easiest action gets done most.

So it’s a combination of great content and great design that drove the amazing growth on YouTube. The design of the site was crucial when the demand finally hit. Without the demand, though, the design doesn’t do much.

In many cases, however, the majority of energy on viral social web apps is placed at launch time. So much energy and focus is spent that it is easy to overlook the tipping points that have catapulted others to glory. YouTube would definitely have been successful on some level, but without Lazy Sunday they wouldn’t have had such explosive growth.

Elephant Math

Seth Godin calls our expectations of viral growth Elephant Math, referring to some calculations Darwin did on elephant breeding. The idea is that when we think of viral, we think of unrestrained, exponential growth. We imagine 1 person telling 10 others, and then those 10 each telling 10 others, and so on.

Seth says:

“It’s pretty clear it doesn’t work that way. Perfect viral growth, even slow viral growth, rarely happens. If it did, we’d have an elephant problem.

The same thing happens with your idea. If one person told four and the cycle repeated itself for a few generations, everyone would know about it. But they don’t.”

The reality is far different, as Seth points out. Viral growth isn’t exponential, it looks more like big spikes, like what Lazy Sunday did for YouTube.

(Source of 83% statistic: New York Magazine story on Andy Samberg – stat is at bottom of page 3)

Update: read Part III

5) Not Appointing a Full-time Community Manager

No matter how prescient your designers and how well thought out your design strategy, there is no way to design a perfect social web site that doesn’t need ongoing management. Yet, some social start-ups fail to recognize this and launch their app without a designated caretaker. The result is a slow failure…the worst kind of failure because it’s not immediately apparent that it’s happening.

In any decent social app, use and users are always changing, always adapting and pushing the limits of your software. So as Matt Haughey, founder of Metafilter, says in his excellent Community Tips for 2007, “Moderation is a full-time job”.

The success of many social start-ups proves this to be true. Flickr co-founder Stewart Butterfield, when asked about making online communities work, admitted there is no silver bullet, but added:

“A lot of our success came from George (Oates), the lead designer, and Caterina (Fake). Both of them spent a lot of time in the early days greeting individual users as they came in, encouraging them and leaving comments on their photos. There was a lot of dialogue between the people who were developing Flickr and their users to get feedback on how they wanted Flickr to develop. That interaction made the initial community very strong and then that seed was there for new people who joined to make the community experience strong for them too.”

Stewart’s description is exactly how George described it to me when I met her at SXSW. She could not over-emphasize the value of her and Caterina spending so much time with users…24 hours a day greeting them, showing them how to use Flickr, and generally saying “Hi”. It was clear to her that a huge part of the early success of Flickr resulted from that personal attention, that personal connection that someone on the other end cares about what’s going on. A full-time community manager is crucial to providing this level of attention.

6) Not Building Archived Knowledge

When your social app begins to grow and you start to attract more and more new people to the fold, you begin to see trends in their initial confrontation with the software. The same issues crop up repeatedly. People have the same problems over and over again and the community manager spends more and more time answering the same questions.

For example, uploading that first batch of photos might be intimidating for those folks who have never done it before. Let’s imagine they all run into the same problem: how do you get photos out of iPhoto and into your Flickr account? There are certain steps to do this, but it is not entirely clear, especially if you’ve never had to export pictures out of iPhoto before.

It’s the community manager’s role to help people at this stage. They’ll chat and email with the person to help them along. But their role should also include figuring out when archiving common problems will make a big difference to a large group of users. If the process of exporting from iPhoto is archived at a URL, then the community manager only has to point people to the brand new “exporting from iPhoto” page instead of explaining it over and over again.

One strategy to avoid repeating the same things over and over again is to use these interactions to feed a FAQ or a user’s guide. Whenever you start to see trends in help, add it to your FAQ and add a section to the user’s guide. This will allow the community manager to focus on the latest, more unique problems without having to rehash older issues again and again.

This seems pretty obvious now that we’ve talked about a general case. But it’s not so obvious when you’re in the heat of battle and these issues are cropping up unstructured for the first time. The secret is to observe patterns in the questions people ask but also in the underlying cause of the questions while leaving enough design time dedicated to creating a healthy set of resources that can serve future users.

7) An Over-Focus on Social Value

This may sound counter-intuitive, but it is possible to focus too much on social value when creating social web applications. Why is that? Well, because much of the motivation within social sites is actually rooted in personal value, or answering the question: “what’s in it for me?”. I’ve dubbed this the Del.icio.us Lesson because it was Del.icio.us who gained so much attention for the social value of tagging but it was really the personal value of saving bookmarks that drove the site.

At the beginning, when you’re building the service, is not the time to focus on social value. There is no social value because there is no user base. So adding tags in the hopes that people will discover new things is probably premature at this stage, for example. Instead, focus on how a single person can use your service even if others don’t share or tag anything.

Think about YouTube, a killer social app. Even at the very beginning YouTube was providing personal value: hosting your videos for free. If they had been charging for this feature, no social design in the world could have caused the growth that free video hosting did. So while YouTube excels at getting viral growth out of the sharing of videos, they’re providing a valuable, personal service at the same time.

It should also be noted that altruistic people, or people who do things for the good of the group regardless of personal benefit, are incredibly rare. They’re so rare, in fact, that they make a very poor population to design for. There just aren’t enough of them to make up a significant population in any area. Even Wikipedians, who have been called altruistic at times, are mostly driven by reputation…the reputation they gain from their peers and other Wikipedians.

Continue to Part III