How big is the blogosphere ?

by John Q on October 2, 2004

And why should we care? I’ll leave this question for later and take a look at some numbers

There have been quite a few attempts to measure the growth of blogging. As this site devoted to the topic notes, Technorati passed its 4 millionth blog a week ago. Both Blogger and Livejournal claim over 1.5 million users, and a broadly similar estimate can be obtained if we take this Pew Study from 2003 and make the reasonable assumption that numbers are doubling annually.

But these are almost certainly overestimates.

There are heaps of dead blogs out there (thanks to changes in technology, hosting problems and so on, I’m already on my fourth). I prefer to start with another Technorati estimate, that there are about 275,000 posts daily. If you suppose (fairly arbitrarily, but consistent with the Pew data) that the average active blog has one post every three days, that would make around 800 000 blogs.

Another way of looking at things is to consider the labour input going into blogs. If you suppose that the average blog post takes an hour to prepare (this includes overheads like research, if any, site maintenance, responses to comments and so on), 275,000 posts per day amounts to about 100 million hours per year, equivalent to the work of around 50 000 full-time, full year workers. As a comparison, the US Bureau of Labor Statistics reports that news analysts, reporters, and correspondents held about 66,000 jobs in 2002.

There’s not much direct financial return to all this effort. There are a handful of bloggers who’ve been employed as journalists, and vice versa. Add in all the tip jars, Amazon referrals, blogads and so on, and there might be enough money to pay wages for a hundred people, spread across many thousands of tiny payments. Turning to blogging software, I don’t know how many people are employed by Blogger, Six Apart and so on, but it can’t be very many[1].

I suspect though, that the economic impact of activities like blogging is significant and growing. Throughout the history of the Internet, most of the innovation has come as a by-product of efforts to facilitate communication within social groups of various kinds (academics, bloggers, peer-to-peer file sharing), rather than as the result of profit-oriented investment. Rather than taking the lead, the business and government sectors have adopted innovations developed in Internet communities, and realised significant productivity gains as a result.

An economy in which innovation is, to a significant extent, a by-product of activities associated with the creation of social capital will have very different properties from those traditionally considered by economists. Some of these issues have been discussed in terms of notions like “gift exchange” but, for reasons I hope to develop in later posts, I don’t think this provides an adequate account of what’s going on here.

1. To get a complete analysis, it would be necessary to look at web hosting services, ISPs and so on, but imputing a share of this sector to blogs would be an impossible task.

{ 14 comments }

1

eszter 10.02.04 at 6:57 am

I think it may be worth distinguishing between different blogs. After all, why are you asking about blogs versus personal homepages or mailing lists or hours spent writing messages in chat rooms? My point is not to pass value judgement about different blogs types, but simply to acknowledge that there really are very different blog types and it may be worth specifying which ones you are talking about. (Depending on where you are going, some may be more relevant to your questions than others.) I (nor others, I suspect) don’t have any data on the percentage of blogs that are very personal journals read by no more than a small handful of existing friends of the author. Are those equally relevant to the questions you are exploring as blogs that talk about arts and culture, technology, politics, economics, etc and which get numerous readers (beyond people who already know the author)? Maybe they are all equally relevant, again, it depends.

I think any definition about blogs, bloggers or the blogosphere is difficult so I understand why you didn’t define these up front, but it may be worth thinking about depending on where you’re going with all this.

Another issue I thought I would raise is whether it’s also relevant to look at _readers_ in all this. How big is the blogosphere in terms of readership? Again, we don’t have much data to go on. And my point is not to argue that if the readership is small it means the whole enterprise is irrelevant, not at all. Nonetheless, it’s another question worth thinking about when pondering questions about the blogosphere (possibly depending on the questions you are pondering;).

2

John Quiggin 10.02.04 at 7:48 am

I agree with all the above, Eszter. The Pew survey had some info on blog readers but I’m pretty dubious about it. In 2003, when the survey was taken, I think most Internet users weren’t aware of blogs. I can remember, after I “discovered” blogging, realising that a number of sites I’d previously visited were blogs.

My main interest is in the ideas I develop towards the end; that things like blogs, chat, P2P and so on are developed in a context where economic considerations are largely absent or irrelevant, but a lot of the innovation in the market economy is now arising as a by-product of these things.

3

Cathal Copeland 10.02.04 at 9:40 am

It would also be interesting to have a breakdown of bloggers by age, gender etc.

My impression is that most bloggers are either students or pensioners, and that hardly more than 5% are female (nerd factor?)

4

Philip Eagle 10.02.04 at 9:58 am

Hi, don’t usually comment (not much of a political blogger) but I think I can be of use.

If they are including LiveJournal in the blog list, then the number of actual bloggers will be hugely overestimated. LiveJournal tends to be based on purely social blogs, media and sf/fantasy fandom culture, and creative art. Large numbers of LiveJournal users have multiple accounts to divide between, say, personal stuff and fiction. There’s also a tendency to create short-lived LJ accounts for pure comedy value.

I will also say that the general impression on LJ is that LJ is significantly dominated by women, and all my personal experience tends to support that. The community functionality tends to give it a bit more of a touchy-feely approach than the other blogging services.

5

marek 10.02.04 at 12:14 pm

I agree with Eszter – readership is a critical measure, not just in volume but awareness. Those within the blogospehere often fail to recognise how hermetic it is. Even now, I keep coming across people who are intelligent, curious, web-literate – but who not only don’t read (let alone write) blogs, but have never heard the word, or if they have, have no idea what it means.

6

Cathal Copeland 10.02.04 at 1:04 pm

“Even now, I keep coming across people who are intelligent, curious, web-literate – but who not only don’t read (let alone write) blogs, but have never heard the word, or if they have, have no idea what it means.”

In Britain I think a lot depends on your political leanings.

I’ve just searched the archives of the Daily Telegraph and The Guardian for ‘blog’.

Daily Telegraph: 21
Guardian: 439

A Guardian reader is thus over 20 times more likely to have come across the term (I didn’t check for ‘Joe Blog’, though)

7

Amardeep 10.02.04 at 2:02 pm

On money: some blogs raise money for causes (especially political campaigns). Isn’t that economically significant? I suspect the money raised by the big left-leaning bloggers alone might be statistically significant in terms of economic ‘output’.

I second (or third) the point about different kinds of blogs.
Some blogs seem close to Internet Discussion Groups (IDGs) — the BBS kind of internet sociality.

The kind of blog that is a soapbox for the author’s ideas, rants, and occasional confessions (the ideal blogger) is a rather different species.

What would be gained by studying the number of blogs qua blogs? Blogging is just a form of web-publishing. It would probably more interesting to study the changing behaviors of individuals (and groups) that write and read blogs. How do people find blogs? What makes them read? How does blog-reading/writing affect other spheres of their lives?

8

eszter 10.02.04 at 2:50 pm

I figured I may not have been addressing the central theme of your post. I think I may have misread the focus of the post’s heading. You are right that users often find ways to incorporate new technologies into their lives that end up being more important and potentially lucrative than the top-down approaches of profit-making companies. There are historical examples of this for other communication technologies as well (e.g. social functions of the telephone that was at first marketed as primarily a business tool).

9

jam 10.02.04 at 2:57 pm

Readership should be reasonably simple to estimate. There’s evidence that weblog usage follows a Zipf distribution. Take the Technorati top couple of dozen linked weblogs. Find their readership. Fit that to a Zipf distribution that goes out to, say, a million weblogs. Sum.

10

Henry 10.02.04 at 3:23 pm

Not as easy as that jam, unfortunately. Kottke’s analysis only covers the top 100 weblogs – there is a fair chance that the distribution does not conform to a power law across all its range, and in any event the r-squared test may not be a very good test of whether or not a distribution is in fact a power law (Cosma Shalizi has a very nice paper coming out that covers this set of issues). NB also that a power law is not necessarily a Zipf distribution (Kottke doesn’t give us the nec. information as far as I can see to determine whether this is indeed a Zipf distribution).

11

jam 10.02.04 at 4:33 pm

Henry: Yes. All true. It may well not be a power law. If it is it may not be a Zipf. For what it’s worth, my SWAG is that it probably deviates downwards somewhere towards the top of the distribution (webloggers dropping out because their bandwidth costs are becoming uncontrollable, yet they can’t attract revenue) and drops off to effectively zero at some point towards the bottom.

But at this point, we don’t need a perfect number. John is, I think, asking for a rough order of magnitude. The estimates in his fourth and fifth paragraphs (number of writers and effort they put in) are clearly not exact. The parallel calculations for readers need not be terribly exact, either.

If Eszter is right that the size of the readership matters, even a ROM is better than nothing.

12

trish 10.02.04 at 4:45 pm

while it would be nice (since I’d expect it to increase quality) if most blog posts took an hour to prepare, I suspect that’s a generous overestimate.

13

Leila 10.02.04 at 7:24 pm

trish: I read that the hour spent on blogs is *total* time per day, not per post. Say it takes two or even three hours of tweaking to set up a blog originally – okay you can get it going in five minutes, but most people tweak, upload files, set up categories, do fancier things I can’t even think of.

Then on a typical day one might read the news and scan whatever sites one finds of interest (research) before composing a couple of posts.

Some bloggers write long thoughtful essays. Some write short blips. Some post every day and often, some post intermittently. Then there’s commenting, on your own blog and on others’. An average of an hour a day for all these activities together doesn’t sound out of line to me.

But an hour per post, of course that’s not right.

14

John Quiggin 10.02.04 at 8:10 pm

Leila, I estimated one post per three days, so an hour per post translates to 20 minutes per day, counting all the things you mention. I think this is probably conservative, but I don’t have a good feel for the blog-related activities of people who post, say, once a week.

Comments on this entry are closed.