Some Thoughts On Popularity

Submitted by Blake on Fri, 12/10/2004 - 20:59

So Arnie asked, "am I being too easily impressed?" [with the LISNews stats]... My first reaction was "yes, you are, we ain't nutin" we are but a little blip on the radar of the internet. But, then I got to thinking. I have no idea just how impressive those numbers really are, maybe they really are impressive? Do I compare us to Slashdot, Metafilter, Fark and BoingBoing? Or do I compare us to The Resource Shelf, Library Stuff, The Shifted Librarian and Librarian.net? Probably more the latter than the former, but I'm going to set out to find where LISNews ranks when compared to different web sites.

First, we already know web stats are notoriously uneven, and inconsistent, but they're all we have. I'll do my best to compare apples to apples here.

So where does LISNews fall in the great blogosphere popularity bell curve? Somewhere in the middle or so…

When we can, let's just consider hits, pageviews and sessions when we talk about popularity from the server side. I'll also delve a bit into RSS feeds as well, but that's a big can o' worms subject to a entire new set of biases and reporting errors. Unfortunately for this piece, there is no single Blogostats page where one can browse stats from any blog of interest. But, luckily, there are some ways we can get some educated guesses at who's doing what in terms of numbers. From what I could find, The Truth Laid Bear: Traffic Ranking Page is about the best we can do. It's an impressive collection of numbers from 250 blogs covering a wide range of topics. [See also a good Look At Those Numbers] Another good piece on this topic is Shirky's Power Laws, Weblogs, and Inequality that goes into some issues with ranking blogs against each other.

So if LISNews was on The Truth Laid Bear: Traffic Ranking Page we'd come in around #40 if we use this months numbers (so far). My only quarrel with those numbers is he's using "visits" a notoriously hard to measure number. What I'd like to do is figure out a nice algorithm that would take into account visits, pageviews and hits, and rank those sites on that magical number. We don't really fit their "ecosystem" numbers, so not much point in trying to put us on that list. So, 40 out of 250 ain't bad! So what other sites play the numbers game that we can try to fit our self into?

Top 100 Technorati has a neat way of doing things, they meausre rank blogs by the number of blogs liking in, kind of a google pagerank for blogs. Not suprisingly LISNews doesn't show up on that list. It's probably no surprise that librarian stories aren't of burning interst to the world at large.

There's a few other sites out there that pretty much replicate this model, or measure something else close to that. Truth be told, it's a hard thing to measure, and they all do it a bit different, and they all do a pretty good job. Jenny is the only one that I've seen on any of those lists. What else can be measured? RSS Feeds!

Bloglines Most Popular Feeds is a good place to start, and it's also a place where more than a few librarian sites show up. Jenny shows up as the #1 librarian feed on the list, with the raltively new LII feed a close second, followed by Research Buzz, and Steven M. Cohen. They all fall in the top half of the list, I think, why would someone have a top 100 list with no numbers?? In any case, this is a librarian friendly list LISNews doesn't appear on. Again, there are several othe sires with similar lists, bloglines seems to have the most librarians on their top xxx list for some reason.

So, to summarize what I've learned so far, The Shifted Librarian is the most poular librarian blog, according to my interpreation of browsing a bunch of sites that attempt to measure popularity. (Since I host The Shifted Librarian I can also say she is the most popular target for spammers as well, probably not the kind of award anyone wants to get!) I think saying that Jenny is most popular is a) an honor and b)correct . It's an honor because she's popular in an ubelievably huge and crowded field full of talented people. It's correct because, well, the lists that rank these kinds of thigs say it's so. But, c) it's not entirely accurate, because these lists are rather insular and limited in scope, and they probably miss many good sites. Special niche sites like librarian.net, and LISNews don't show up here because we are popular in a different way. We have limited appeal to a limited audience and aren't linked to by the blogosphere because what we write about isn't very interesting to most people. Even though these numbers are far from perfect, they're all we have, and they all pretty much agree "Shifty" takes lithe LISBlogoprize for most popular librarian blog. She's got a great domain name, great posts that appeal to a wide variety of folks, and a great hook.

So what ways can we measure popularity? By my way of thinking the broadest measures are the best. That's why I say Jenny is the most popular. She shows up on the most lists, and since I can cheat and look at her site stats on LISHost, I know her site is quite popular. But because of the nature of the different ways we measure such a diverse number of sites and places there are many different factors to keep in mind. Other things to think about [that we can't measure very well]:
Number of folks on the mailing list (from what I hear resourceshelf would win hands down).
The afformentioned Apache logs, as best I can tell LISNews comes out on top there on some measures, but is beat by Shifted in others.
Number of subcribers to RSS Feeds, that's a Jenny victory again.
What about number of participants and authors? Not a fair comparision, but nonetheless one that interests me, since that's what LISNews is all about.
Number of links between the library blogs. If I had the time I could pull those numbers out of LISFeeds.
The diversity of pages, that is, how many different pages are being viewed and hit.
How frequently the site is updated, how long it's been around, and so on.
All interesting things that would be fun to study.

What's important to me about the LISNews numbers is not necessiarily the raw numbers, but that we have a wide range of people participating, reading, and sharing stories.

So to answer you Arnie, no, you're not easily impressed, I'd say our numbers are impressive! Having watched them grow from 0 to what they are today I'm mightly impressed that so many people see some value in what we're doing. There's probably only one other thing I've put more time into than LISNews, and that's the deck I built this summer on my house, not something easily shared with people around the world like LISNews is!

It's impressive that there is a wide range of people reading a wide range of stories and participating in someway on the site, and yes, the numbers themselves are impressive, even when compared with most other sites. And you should also be impressed with The Shifted Librarian's popularity!