I did an experiment this morning.  I analyzed my blog traffic/views using all the data since December 2003. For some reason I haven't done this before.

I wanted to see how closely my page views traffic looked like the Long Tail.  I got more than I bargained for.

This is roughly what the classic Long Tail shape looks like:


Figure 1, The classic Long Tail

My Long Tail

I took the page view data of all 535 posts and sorted by page views (PVs - over 3.4 million total since December 2003) in ascending order.  This is what I saw:

Figure 2, My classic Long Tail by PVs

And there it is, a Long Tail if ever I've seen one - a small number of posts taking a disproportionately large percentage of the total page views, curving off dramatically with a gradual tailing off and the rest getting fringe-type attention. The top 10% of the posts equate to 24% of total PVs. The top 20% of the posts take up 36% of total PVs.

Quite uncanny, almost to the point of obviousness in this mornings prediction. But it could have been different. In fact if I also look at the RSS views (see below) the Tail could be thinner or thicker or longer or shorter.

PVs of Latency

The following chart sorts the posts by date and shows PVs for each post, the earliest first (Dec 2003 on the left) and around 22 months if data at the time of writing this. Now at first sight it might appear that my blog is getting worse over time as the PVs in the latter 7 month are roughly averaging around half the of the first 15 months.  In fact what we're seeing is the effect time has on post PVs.  Today I'm getting more PVs in the first week of a post being live than I did a year ago, at least 6 times (I record my monthly stats at the end of each month). So at least half the volume of the PVs on the chart were served 15 months after they were posted. Latency in action.

Figure 3, PVs over time, earliest post on left

Where are the PVs coming from? Search engines, Google in the main.  At least 80% of the traffic I get to posts after 3 months are via the search engines that match niche content with niche interests - the Filters of the web.  The rest of the traffic comes after 3 months, from referrers from other blogs and online resources (articles, guides, lists of useful links on a subject, etc.).

The RSS Effect

Here is the same chart but with RSS views overlaid (Figure 4).  RSS views are the number of times a post (item) is viewed via RSS (obviously).  Some of the more immediate PVs after posting will be click-thrus from the RSS feed for about 2 weeks after the post has gone live, but I suspect there is some double-counting here in Unique Users (UUs - not shown) as one user might see both a post on the blog and in a reader/aggregator.

Total RSS views is 3.2 million (red line), almost exactly matching the 3.4m PVs.  Notice the long term RSS views trend/average (up - black lines) is inverted compared what we saw with PVs (down, but will go up over time)

RSS has effectively doubled my 'traffic'.

Figure 4, PVs (blue) and RSS view (red line and black line average) over time, earliest post left.

In the first half of the chart (the first 11 months), RSS views pretty much track PVs, but you can see the delta on the second half of the chart: RSS is double the views of PVs. It'll be interesting to see how this trend develops over the next year, but it seems clear that RSS is making its impact on my blog use at least.

RSS searching and filtering services such as Technorati, Ice Rocket, and most recently MSN will make a greater and greater impact on traffic patterns as their usage increases.  As we've seen with traditional web search, the traditional web UI is no longer primary interface to much of the content on the web. Users are finding content, not browsing for it. They are beginning to subscribe not 'visit'.  Navigating the web is becoming a less structured affair (of course,  IAs, site maps and navigational systems - traditional interfaces - will always be with us, for a while anyway).

My blog as it turns out has a Long Tail, and that is not a bad thing - people are finding the content as they need it.  But if I were in the business of 'moneterizing' my content (my spell checker has just had kittens) I'd want to thicken and shorten the tail, increase volume of PVs, increase distribution and get better returns on the investments made for content development / acquisition.

The following is a comparison of three possible charts of my blogs PVs (Figure 5). The bottom left 'Actual' is the PVs data we looked at above in Figure 2, this is fairly classic Long Tail shape...I'm rather proud of it :-).  Stating the obvious here somewhat, but what I'd really want is the curve to look like the 'Web 2.0 ultra effect' curve on the far right- squashed upwards with a Short Tail and as thick as possible.  This in fact is a Long Tail rotated 80d anti-clockwise, then mirror-flipped and thickened with volume. Very little wastage on terms of utilization.

Figure 5, The Long Tail and the Web 2.0 Effect

The middle chart above 'Web 2.0 effect' (Figure 5 still) is closer to what Web 2.0 can achieve, and has in my case due to RSS.

Below is my Long Thick Tail charted as stacking RSS views on top of PVs (Figure 6):

Figure 6, The Long Thick Tail, thickened by RSS

This is no longer a classic Long Tail -  it is thicker and RSS constitutes half the total volume of all views.  This is the Long Tail with Web 2.0 effect or The Long Thick Tail.  The content on my blog has been set free by RSS and the feed search engines, unconstrained by the tedium of a navigation system.  These Web 2.0 technologies are amplifying the value of my content by getting more eyeballs to it.  They are literally creating new value.

On this subject, Joshua Porter, in a short but sweet post (unlike this one!) that has propelled my own thinking on the connected concepts of the Long Tail and Web 2.0. It is why I did the analysis, so thank you Josh. He puts it this way:

"I see lots of similarities between the Long Tail and Web 2.0. Both ideas are about improved access to previously unavailable content, both are about showing the whole catalog, and both are ultimately great at enabling user choice. They seem to overlap a lot. If I had to make a marked distinction between them I would say that Web 2.0 is about the access to information while the Long Tail is about the economics of it all."


Opening Up

The BBC get this connection, the relationship between the Long Tail and Web 2.0. At least they say they do although they've not released any APIs yet, they understand what's coming and are preparing.  They are preparing for the remixed, mashed up, open and recombinant web enabled by web services, APIs, and RSS (buzzword bingo Alex, 10 out of 10).

Figure 7 is my attempt to visualize all this.  Content gets created, stored and made available through traditional UIs and then freed through RSS, web services/APIs for others to amplify and create new value, who in turn create more UIs, more RSS feeds, web services and APIs and value, and so on.

Figure 7, Web 2.0 enabled content via services/APIs & RSS creates amplification and new value

"Power in the Web 2.0 comes not from controlling the whole system, but in controlling the connections in a larger network of systems. It is the power of those who create not open systems, but semi-open systems, the power of API writers, network builders and standards definers."

I think William Blaze is right here about how open a system can be without diminishing the value it is there to create.  So while I partially agree with Richard MacManus' philosophy of 'letting go of control and freeing your data' (to paraphrase) it comes down to staying in touch with reality (not just capitalist reality this applies to the gift and attention economies too )...any philosophy that fails to take account of reality becomes academic at best and not practicable.

I suggest a working philosophy is really 'let go of enough control and free most of the data'.

Web 2.0, the Long Tail and Business Models

An example of this philosophy at work is eBay.  There are certain parts of its business (process & content/data) that will need to remain proprietary and 'own' for it to remain competitive. An obvious example is the customer data it has acquired. This is an ultra-high value asset to eBay (of course the privacy law constraints are another good reason for not opening this data up).  This is the kind of data it wants to keep under lock and key, not open.  The data it does want to and does share are category listings, product details, pricing data, bidding counts, etc., that make up the core of the value eBay has to offer third parties.  Those third parties can then create new interfaces (UIs and programmatic), recombined with other data to create new value.

The user interface on eBay.com is the primary customer interface for its 56 million customers, but is one of many used to interact with the powerhouse.  These other interfaces are distributed under its Affiliate program, including the APIs and services (such as Paid Search).  This is the Web 2.0 enablement of  eBay's content, data and services and helping  make their Long Tail thicker.  The Long Tail in this case being the 10,000 affiliates of  who participate in the distributed ecommerce program and growing revenues along with eBay through a variety of business models.

Now, there's nothing new here... the pure-play does this instinctively. This type of distributed business model is what it means to be a successful ecommerce business and eBay, Amazon and others have been doing it for years.

When the Long Tail and Web 2.0 meets Flat World economics we get players to emerge like AllofMP3.com, a Russian online music service providing pretty much any file type you like of pretty any much any music out there.   The service has already caught the attention of the RIAA and I'm not sure how long they might last.  At $1.50 a CD (you pay by the meg, if you want high bit rate, you pay more, like 50c more) and DRM-less you're playing by Russia's economics where CDs cost $3. User imported digitally, cheap but legal.  As well as the web UI, there is a download client with searchable local catalog database and with integrated web UI for payments and account management transactions.  I have no idea how much they are selling but I'm pretty sure they are raking it in.  I'm actually spending more on CDs than ever, I don't think I'm alone in thickening up their Long Tail.

Again nothing new, been around a while.

What is new are the gazillion remix opportunities that are emerging for everyone else, not just the well funded players. Through insights provided by people like Chris Anderson on the nature of the network economy, the Long Tail economic theory throws new light on old problems and combined with the direction that the web is going in - known as Web 2.0, there are some very interesting and good times ahead.

(PS, Let me know if you've got Long Tails of your own to share, am especially interested in RSS-related Tails, thanks.)

(Trackbacks are knackered again. Here are some blogs pointing to this post).