Common Pitfalls of Building Social Web Applications and How to Avoid Them, Part I

Originally published on Josh's blog Bokardo on June 5, 2007.

In the last several years we’ve seen the rise and fall of many social web applications. While most of our attention gets paid to the hugely successful ones like YouTube and Facebook, we can also learn a lot from those that have failed. Here are some of the common pitfalls that lead to failure when building social web applications.

1) Underestimating The Cold Start Problem

If you build and release your social web site and nobody uses it, you have the cold start problem. This problem affects most social sites, and directly results from designing for the network. The effect of the network is that nodes on the network (web sites) have attention momentum. We pay attention to certain nodes (sites) already, and so if you’re trying to add one to the network then you have to build your own attention momentum over time. This is not easy.

Too often, though, this hurdle is underestimated. The first step is to admit there’s a problem. Say “This is not working. Our early users are not using the site how we want them to.” You would be surprised at how often this doesn’t happen. Instead, what often happens is that more money is pushed into features or marketing, which is precisely the wrong move.

Strong social sites build value one user at a time. If one user finds value, then they’re much more likely to tell others or invite their friends. Strong sites don’t succeed by attracting “markets,” satisfying entire groups of people with a certain feature set. Instead, they succeed on a smaller level, really focusing on individuals and their immediate social network. Then they can branch outward. One strategy in particular is to design for your friends, get the system working well for them, and then release it to a broader audience.

2) Focusing On Too Many Things

I got this email in my inbox the other day from a well-meaning entrepreneur who was building a new social web site:

“(our site) aims to combine the best elements of Digg, Del.icio.us and StumbleUpon, as a mechanism of social discovery and personal expression—but with the unique element of real-time.”

I get so many of these it’s not funny. This is a clear case of focusing on too many things. If you can’t describe what your site does with a single, clear idea then you’re trying to do too much. In addition, a comparison to other sites in this way is a bad idea, because they’ve already beat you. They already have a strong brand while you have a weak one.

The ease of adding social features makes overload likely. Development frameworks make adding friends, tags, profiles, blogs, or a host of other social features much easier than it was even a couple years ago. This is the opposite to a barrier to entry, where the hard part is building something at all. Instead, the ease of adding social features is a barrier to focus. If you have every feature under the sun you’re probably not focused as well as you could be.

So, focus on one thing that isn’t being addressed. It can’t be something like “the unique element of real-time”. It has to be something inherently valuable, like a common frustrating activity. Nail that one thing to the ground, and show people how you do that one thing better than anybody else.

Think of the most successful social sites out there. They usually focus on a single thing. YouTube (video), Netflix (movies), eBay (auctions), MySpace (friends), Flickr (photos), Del.icio.us (bookmarks) and most of the social features on those sites are aimed at making that one activity better. These are just the giants. There are many more niches that are successfully designed that are even more focused: for example, threadless focuses on t-shirts, Last.fm on music, etc…

3) Lack of Sustained Execution

What makes Google so terrifying to their competitors is that they never stop getting better. They’re executing each and every day to make their software the best it can be. For example, in September of last year they did the unthinkable: they completely killed off the interface paradigm of a solid, growing product, their Google Reader software. But they replaced it with an even better interface that was universally acclaimed.

It’s too easy to fall into the desktop software mindset of build, release, and wait for the next cycle. But with social software, you don’t have the opportunity to stop improving. Your community is always growing and changing and so your management has to as well. There will always be things to do, screens to improve, questions to answer, wording to tweak, and support docs to update.

This can seem daunting, but it’s mostly about mindset. If you see it as a sustained problem, then it will be one. If you see it as an opportunity for continual improvement, your outlook will be more positive.

4) Pointing the Finger When Missteps Happen

When you mess up on a social web app, as you undoubtedly will, you have to come completely clean or your users will smell your fear and hate you for it. Social sites are not typical software… they ebb and flow depending on the community and how it evolves over time. You, as the manager of a community, must act accordingly.

Consider the recent Digg dustup in which the Digg community pushed back on the site after they tried to remove a certain DVD-cracking code from user-submitted entries. At first, Digg tried to explain the situation away by saying they were legally obligated to as the result of a cease-and-desist letter. The basic message was “our hands are tied.”

But then the Digg community overwhelmed the site and got the DVD crack code up anyway. The failure of Digg management to stand up for their users initially resulted in the user’s aggregate behavior. Digg didn’t lose out, however, as this community passion provided an opportunity for them to ride the wave, so to speak, reversing their course and standing up to the cease-and-desist. Their apology letter and reversal suggests they quickly realized that pointing the finger wasn’t the right course. Only by accepting responsibility for their user base could Digg keep their respect.

Here’s a template for how to say you’re sorry.