20 Comments

Summary:

Chris Anderson in his book, The Long Tail (read Chris’ book), divides the power law distribution curve into only two segments… the hit-driven head (Big Head) and, obviously, the long tail. What’s missing is actually the most important part… the section in the middle of the […]

Chris Anderson in his book, The Long Tail (read Chris’ book), divides the power law distribution curve into only two segments… the hit-driven head (Big Head) and, obviously, the long tail. What’s missing is actually the most important part… the section in the middle of the curve The Fat Belly.

It has implications for social networks, and other communities. Take a look at Digg’s technology section. All the articles in the Big Head received about 250,000 votes in total vs. estimated 2.5 million votes for the ones in the Long Tail. As for the Fat Belly, those stories got a whopping 10 million votes! Now that’s what I call a healthy “middle class.”


Many of you will recall the big brouhaha that erupted when Jason Calacanis re-launched Netscape.com as a Digg clone… which I was reminded of when I read this recent article about Netscape in the Washington Post. What was controversial about Calacanis, of course, was his announcement that he would poach talent from his competitors by paying, for the first time, the elite “social bookmarkers” of the various socially-curated news sites.

Privately, I applauded Jason for having the courage to experiment with new business models. Jason, shrewdly, wanted to spur the growth of Netscape by leveraging the simple fact that a small group of top contributors accounted for an inordinate amount of influence when it came to the process of determining which articles/stories would become popular (e.g. stories that had been “Dugg” to the front page). Evidence to support his strategy surfaced when it was revealed that, in fact, the top 10 Diggers, from a population of approximately 450,000 registered Digg users, were responsible for an astonishing 30% of frontpage stories (alternatively, the top 100 Diggers were responsible for over 55% of frontpage stories). To me, Jason’s strategy represented an interesting derivative (call it a web 2.0 version, if you will) of supply-side, trickle-down economic theory/policy. But even while I applauded Jason for his boldness, I was very skeptical that his strategy would work. And here’s why.

Having been involved with the Internet since literally day zero of its commercial inception (back in November 1992), I have had a front-row seat watching the power of our medium’s “democratizing” effects. Digg is a great example of such effects as the value proposition of the entire system is reliant on the most democratic of all instruments… one person, one vote. Consequently, I’m a big fan of Chris Anderson’s “Long Tail” thesis, as I find the framework to be remarkably consistent with my own experiences and observations. Yet, when one attempts to understand the strategic implications of Digg, and its role as a democratizing force, even the Long Tail paradigm fails to fully showcase its true, disruptive nature.

While I won’t go into the details of the Long Tail thesis here (read Chris’ book), suffice it to say that what I find limiting is the fact that he divides the power law distribution curve into only two segments… the hit-driven head (what I call the “Big Head”) and, obviously, the long tail. What’s missing, in my opinion, is actually the most important part… the section in the middle of the curve that I affectionately call the “Fat Belly”.

The recognition of the existence of the Fat Belly is critical for many reasons, but allow me boil it all down to this overarching statement: Any economist or political scientist will agree that the health of any democratic society that’s fueled by free market capitalism is measured by the robustness of its middle class. A large and vibrant middle class demonstrates a healthy redistribution of wealth within a nation and its economy, ultimately serving as a catalyst for the power of one vote and equality amongst its peers/citizens. What all this means, and to bring this back down to earth, is simply that I prefer to segment the power law curve into three distinct segments… the big head, the fat belly, and the long tail.

Now, why is this relevant/important? The answer is simple… in my view, the potential success of any Internet venture, particularly for those heavily reliant on the development of an online community of active participants, is directly correlated with the concept’s ability to create a large and dominant Fat Belly… much like a successful democracy will result in a large and dominant middle class. And towards such ends, Digg is one of the best examples I’ve seen. But be warned… there are subtle forces at play within an online community like Digg, some of which are seemingly contradictory, that could easily lead observers and decision-makers astray. For instance, how can I say that Digg is a truly democratic community, given the fact that a small group of Diggers have such disproportional influence within the system? To answer that, let me step back a little so I may frame my argument.

As I’ve said previously, if Hollywood was a country, it would be a nation with no middle class… you’re either a superstar producing blockbusters, or you’re a waiter. The scarcity of distribution and production that drives Hollywood guarantees such inequity and extreme division. Contrast such constrained dynamics to Digg… where users can contribute by submitting *any* article from any publication or blog in the world (basically an unlimited supply to choose from) and where any member can vote to influence any given article’s popularity. In other words, the fundamental supply/demand equation behind Digg, driven by the Internet’s mantra of abundance in production and available inventory, sets the critical first stage. But here’s the real proof of how truly democratic Digg is… let’s go empirical.

I did a quick analysis of Digg’s Technology section, which is the longest-running and most active forum. Specifically, I looked at the compilation of all the most popular stories that were Dugg this year. Incidentally, for a funny (yet accurate) explanation of how Digg works, read Nick Douglas’ description in Valleywag.

As expected, the distribution of the approximately 9,300 of the top stories this year charts into a power law curve. The Big Head, which I define as those articles with over 5,000 votes, accounted for a mere 32 articles. The Fat Belly (those with 1,000 to 4,999 votes) had nearly 4,000 articles. Lastly, the Long Tail (less than 1,000 votes) had the remaining 5,300 articles, predictably the largest number. (n.b. noticeable changes in the gradient along the curve served as the primary factor in my choosing the points of demarcation).

But here’s where it gets interesting. All the articles in the Big Head received about 250,000 votes in total vs. what I estimate to be around 2.5 million votes for the ones in the Long Tail. As for the Fat Belly, those stories got a whopping 10 million votes! Now that’s what I call a healthy “middle class”.

So herein lies the conundrum… or seeming paradox. How does one reconcile for the fact that, on the one hand, a very small group of Diggers are responsible for seeding the vast majority of the popular stories, while on the other hand, the end result for the articles themselves is highly democratic in terms of the eventual distribution curve.

For me, the implications are straight-forward. It doesn’t matter *who* initially Dugg the article. Rather, it’s far more important to have a critical mass of subsequent voters for whatever article may appear on the frontpage. After all, at the end of day, the value of Digg is not represented by the existence of an elite club of Diggers with the disproportional ability to influence votes; instead, the value lies in the articles themselves, and how the population ultimately votes for each one. What this also suggests, on the flip side, is that a winning strategy is one that will develop and empower Fat Belly contributions, and any strategy that alienates them puts the entire system at risk.

As stated earlier, I believe Jason’s willingness to experiment with different business models is very admirable. Doing so with his blogging network certainly worked out well. But the community dynamics behind social news sites are very different than those at play within the blogosphere (which is very personality driven), and I fear that any strategy that “castes” an upper class of contributors is antithetical to the core principles that fuel a site like Digg.

Robert Young is a serial entrepreneur who played a major role in the invention & commercialization of the world’s first consumer ISP, Internet advertising (pay-per-click ads), free email, and digital media superdistribution.

  1. Very poignant piece, Robert… and I fully expect “Fat Belly” to be all over Techmeme tomorrow. :)

    The middle class analogy is a great one, and certainly one of the reasons why the argument that some elite set of story-submitters is what makes you great is a faulty one. The answer to why such a small set of story-submitters accounts for such a large amount of “front page stories” comes down to social dynamics much more than anything else. Once you’re an established voice on Digg, Newsvine, Netscape, Reddit, or wherever, you have more people “watching you”. Every time you submit a story, you have an army of supporters ready to back you up and vote for what you’ve voted on. It’s not exactly wisdom of the crowds because the independence aspect is missing, but something slightly different. It’s not even necessarily a bad thing, but that’s how it works… I’ll be speaking to this at Carson’s Future of Web Apps Conference in SF in a couple of weeks, in fact. What it comes down to a mimicking of what happens in real life when fads come into style. A popular person starts wearing Hush Puppies and since he has a group of other people watching and respecting what he does, several other follow… and it snowballs from there. It’s not elitist. It’s just social dynamics.

    Share
  2. Awesome post, thanks! Your “fat belly” theory very much reflects my experience with search terms and traffic. While there is certainly a lot to gain in the long tail, a great set of medium competitive “middle class” keywords produce the most traffic in most niches!

    Share
  3. A big post about distribution curves – and not one diagram to help me visualise ‘the fat belly’. Sorry, but I got bored reading about half way through and moved on ..

    Share
  4. Ummm, this argument seems confused. It flips back and forth between two different statements:

    A) A very small percentage of the userbase generates a very large portion of the stories.

    B) Of those stories, a very small percentage are extremely popular, and others are moderately popular.

    It then seems to say that if you added up the popularity of all the moderately popular stories, it would far exceed the few superpopular stories, therefore – MIDDLE CLASS, etc. etc. However, that’s meaningless.

    Or may it’s that the value of a superstar isn’t in the huge salary he or she commands, it’s in the large number of people willing to pay to see the superstar, therefore being a superstar is really a proof of democracy. Which is quite twisted.

    The theory is that the super-diggers have an above-average ability to create appealing posts, and that’ll hold whether it’s done at digg or netscape. There’s nothing democratic or middle-class about that idea.

    Share
  5. Seems to me you have your bell curves and power laws mixed up. There is no “fat belly” in the graph of a power law. A power law by definition tapers off drastically from the “big head”, while the classic bell curve does indeed have a fat belly.

    Share
  6. Presume you are talking about the slope of the curve not being as steep as the typical Long Tail graph makes it out to be (as someone commented above, a visual represntation would surely have helped)

    That said, I would hazard a guess that this ‘Fat Belly’ concept may also apply to eBay sellers

    A few ‘Power sellers’ with very high feedback selling lots of stuff, followed by the ‘middle class’, folks with medium feedback accounting for the bulk of the sales and then the long tail of low feedback sellers

    Share
  7. Isnt this very similar to Brad Feld’s post couple of days ago on the 80-19-1 rule?

    http://www.feld.com/blog/archives/001937.html

    Share
  8. Thanks for all the comments. Allow me to address a few…

    SimonG… you’re right. I probably should have put up a diagram. And I will include one when I write a follow-up piece that focuses further on my Fat Belly thesis.

    Seth… hope my follow-up clarifies the confusion.

    Greg… while I can see why you may think I have the distribution curves mixed up, I am indeed talking about a power law curve.

    Adam… I actually wrote my piece before Brad Feld’s post (it’s just that Om held onto it for a few days before running it). I just read it, and we do seem to be saying similar things.

    Before I end, let me say one more thing. An example that Chris Anderson likes to use with his theory is the sales distribution of books at Amazon vs. Barnes & Noble, as they relate to bestsellers and the long tail. For me, it was interesting to find out that in the case of B&N, about 5% of sales are derived from bestsellers. The vast majority (more than 60%) of sales is what I would consider Fat Belly (by authors who are relatively well-known, but never quite make it onto the bestseller list). In fact, a key reason their superstores carry up to 200,000 titles is to capture the Fat Belly.

    Share
  9. So, are you saying there is a “belly” on the long tail curve? Or are you imagining a different kind of curve, like this one (which is described as the “big belly”): http://www.b-list.org/weblog/2006/05/30/lets-talk-about-frameworks

    Share
  10. I break user up into the “three c’s” (old school!):

    1. creatives (1%)
    2. contributors (19%)
    3. consumers (80%)

    In the Netscape/reddit/digg world:

    a) The creatives are the ones who put in stories.
    b) The contributors are the ones who add comments and vote.
    c) The consumers are the ones who just ride on the work of the first two groups.

    In order to get the consumers (the 80%) you need the creatives and contributors (the 1% and 19%).

    In order to get the contributors (the 19%) you need the creatives (the 1%).

    In order to get the 1% you need to motivate them somehow.

    At Netscape we motivated them with compensation, recognition, and affiliation.
    At digg they motivated them with recognition and affiliation.

    Netscape is building the 19% right now, and in many ways we are an anomaly since we started with a large amount of traffic due to Netscape’s heritage. We brought in the 1% and they are doing an amazing job. We’ve got the 80% covered, and we right now we’ve been building up–and frankly educating–that 19% on what social news even is.

    People in the web 2.0 bubble forget exactly how niche what we do is. The mass audience doesn’t understand what social news and bookmarking is. It’s gonna take another 2-3 years for them to figure it out. Just like it took 3-4 years for blogging to stick. However, I’ve seen this story before and it comes to resolution quickly.

    Of that 19% we need to get I’d say we already have 1/3rd of them, and by the end of the year we’ll have the other 2/3rds. You can’t build a community overnight, it takes time. I have to say, that I’m thrilled we got to 50,000+ registered members in < 2 months. We’re adding well over 1,000 a day and growing–we’re way ahead of our plan.

    Share

Comments have been disabled for this post