Post1000versary, and skepticism related to growth statistics

Milestone: Post #1000. Is that half a working year in total?

There was recently yet another spate of articles on blog statistics. I remain skeptical of the precise numbers, given that nobody else can examine them, as unverified reports are often wrong. But the interest is a good reason to reflect on what such growth statistics mean (especially since the press eats up the hype, and it'll be echoed many times).

While it's unarguable that there's growth, I think there's some questions as to where the growth is going. My conjecture is that it's going first to increasing numbers of young people chatting with friends (e.g. MySpace), then to generally popular pundits, then a little to local A-list BigHeads, and last of all to the Z-listers. So doubling of the total number of the bogosphere doesn't necessarily translate into doubling to the average blog-writer. It's tricky to establish this, though, because there's definitely an increase in automated retrievals of pages, and that *will* affect everyone to some extent.

It's very important to examine raw data with care. For example, I get some hundreds of image retrievals a day from various piggybackers using my site bandwidth to display icons, something which I haven't bothered about since it's relatively trivial. But if I mistakenly believed that it meant anything, I'm sure it would contribute to an impressive but meaningless number of hits (as in, "I get blah-blah unique IP addresses visiting my site per day").

There's also numbers which do not mean what you might think they mean. One aggregator-maintainer said there were around 200 subscribers to headlines from my blog. But when I checked against my own log files, it seemed that traffic from there was only one or two readers per day. The number was true. It just didn't mean what it sounded like it meant, what would be easy to believe it meant. Note this wasn't a read-by-feed issue. Rather, ~ 200 headline subscribers translated into one or two real readers.

I decided to look at some subscriber statistics compared to about six months ago

So, on that measure, there's been a roughly 20% increase in six months. Or, in absolute terms, a whole whopping *52* subscribers (am I A-list yet?). Not that I turn anyone down ... but it does present a different perspective than breathless bubbleness.

Seth Finkelstein's Infothought blog (Wikipedia, Google, censorware, and an inside view of net-politics) - Syndicate site (subscribe, RSS)