Wednesday, October 29, 2008

The Web Travels in Deep Ruts

Recently I came across an older NY Times article on what is called "cumulative advantage":
Is Justin Timberlake a Product of Cumulative Advantage?
Cumulative advantage is the process by which people, when choosing something -- music to listen to, a product to buy, a movie to watch, a book to read -- are strongly influenced by the choices that other people have already made in the same space. In a nutshell, it is the process by which the rich get richer and the popular get even more so.
"...when people tend to like what other people like, differences in popularity are subject to what is called “cumulative advantage,” or the “rich get richer” effect. This means that if one object happens to be slightly more popular than another at just the right point, it will tend to become more popular still. As a result, even tiny, random fluctuations can blow up, generating potentially enormous long-run differences among even indistinguishable competitors — a phenomenon that is similar in some ways to the famous “butterfly effect” from chaos theory. Thus, if history were to be somehow rerun many times, seemingly identical universes with the same set of competitors and the same overall market tastes would quickly generate different winners: Madonna would have been popular in this world, but in some other version of history, she would be a nobody, and someone we have never heard of would be in her place."
This idea has all sorts of applications. First off, it makes market predictions nearly impossible. You can't predict which product will conquer all others because the intrinsic elements of that product -- its price, its quality, etc. -- are only partially responsible for its eventual success or failure. Take the iPod, for instance. It wasn't the first MP3 player. It arguably wasn't even the best when it first hit the market. But it was good enough for lots of people, and once it seemed like everyone had an iPod, Apple had the sort of thing that marketing money can't buy: the implicit endorsement of an entire population. When you're shopping for an MP3 player, what are you going to buy -- the Sandisk player that seems all right but which you don't know much about, or the iPod that lots of other people have bought before you (and it must be good, otherwise people wouldn't be buying it)? Odds are you'll make the choice that everyone else is making, thereby adding your weight to the iPod's cumulative advantage.

Clearly this is a serious problem for anyone trying to enter an established market, since you have to overcome the cumulative advantage of the rivals who are already there. That's obvious. But I think there's an interesting take on this question for website design: namely that cumulative advantage can also influence your traffic. I'm not just referring to Google's PageRank, though that's certainly a factor; Google weights more heavily sites that are referred to from other sites, so search will give a clear (cumulative) advantage to a site that's already out there, gathering links. But I think it may go deeper than this.

Consider a community-oriented site. Let's say it has articles, news posts, maybe a blog, maybe a forum, maybe even a wiki. Some of that content reflects traffic, some of it doesn't. Forums clearly display where the most posts are; they also often display elsewhere on the site what the most recent posts are. Blogs display traffic by way of comments. Wikis display traffic by the weight and detail of articles. All of these send a message to a site's visitors: this is where other people have gone. So, when you're on the site for the first time and wondering what to click on, where are you likely to go? To the (perhaps excellent) article or news post that (for all you know) has never been read before, or to the parts of the site that you can tell are attracting lots of attention? I think I already know.

It would be great to test this. I'm envisioning an A/B test in which some element of site content -- let's say news -- is published both with and without traffic figures (maybe a little feature below the news title that says, "Article read X times"). I'm willing to bet that showing readers where the traffic before them has gone will strongly influence them to head in that same direction themselves. Of course, this would mean a slow and lingering death for any content branded with the "Article read 0 times" mark of shame. You win some, and you lose some.

Bottom line takeaway: popularity and quality are rarely the same thing, and most likely that functions on the micro level as well as the macro. So don't take it too hard if your best stuff is getting next to no traffic; it may be that something else just happened to gain a cumulative advantage.