The Need for “Network Effects”
If you pay attention to the blogosphere you probably hear about a dozen or more new, potentially useful websites every week. The low cost of entry, the presence of increasingly sophisticated tools and the emerging ubiquity of the web have led hundreds of entrepreneurs to create sites, web applications, mashups, plug-ins and widgets that do all manner of things to make your life interesting.
And that’s great for us as users of the web. The problem for entrepreneurs is that if you come out with something shiny, how do you keep fickle users from migrating to the newer, shinier thing that comes out a month or a year after yours?
There are a few decent answers to this question. The old-fashioned answer is intellectual property (typically patents when you’re talking about technology), which can keep anyone from competing with you, but realistically most entrepreneurs bootstrapping their idea for a social media site aren’t going to file patents (nor would they receive them).
The answer that most “Web 2.0″ companies (a phrase I dislike, but rarely successfully avoid using) come up with is network effects (whether they use that phrase or not). Network effects refer to products (in this case web sites or apps) that get better the more people use them. While a lot of the stickiness of sites with strong networks is associated with the “new web”, probably the best example of all is Ebay. A good team of developers could recreate Ebay in a week or two, but competing with Ebay would be virtually impossible. Ebay has millions of users, EbayClone has zero, and there is significant pain for consumers switching away from a site with enormous activity to one with none. So even if EbayClone had a better interface or other advantages, their ability to compete with Ebay would be pretty low.
LinkedIn is another great example of successful deployment of network effects. What they’re doing isn’t rocket science – it’s basically an online rolodex-builder. But with about nine million users, others that try to compete with them have a high wall to clear before they’ll be successful. If I convince a hundred of my business contacts to connect to me on LinkedIn, it’s going to be a huge pain in the ass to switch to another one and reconnect with all of them. The network gives LinkedIn a barrier to entry even though they have no IP protection and their technology is not novel.
This was recently brought to my attention with New York subway system mashups. For over a year I’ve used This Subway Map to figure out which trains most easily get me where I’m going. But then This One was released and I switched overnight. Why? Gypsy Maps gives me directions – which train to which train from Point A to Point B. Do I really need this feature? Not really – I can look at the subway map and figure it out. But there’s no switching cost for me. No pain associated with ditching OnNYTurf’s map and moving to Gypsymaps, so the slight incremental improvement is enough to induce me to switch.
All of which is a long way of saying, think hard about how you keep your customers. If you are building a site that has no IP and low barriers to entry (cheap to replicate, simple features), you need to figure out some way to keep your customers on your site when something shinier comes along. That could be a social element (their friends are on your site), a personalization element or some kind of hosted data advantage. It could be something else, but you need to have some kind of friction associated with leaving. If you have users, someone out there is trying to take them away as you read this.
