Comments on: List of library blogs

By: tangognat

tangognat — Tue, 16 Aug 2005 03:03:49 +0000

It’ll be interesting to see if there are better tools for measuring things like this in the future. Or ways of figuring out what the data means. For example, I can see that I have had 1,718 hits on my feeds so far this month, but I’m not sure really what that means.

By: Stefan Hayden

Stefan Hayden — Tue, 16 Aug 2005 01:30:58 +0000

I know a lot of RSS reader do show how many visitors but I’ve always had problems with that. While Bloglines is supposed to show how many subscribers you have I’ve found the number to inconsistent with what feedburner shows. While it’s wonderful that some RSS readers show some effort there definitely is some work to be done.

By: tangognat

tangognat — Mon, 15 Aug 2005 22:47:18 +0000

No, I don’t think the survey is worthless without incorporating log file data. I’d be happy to see it repeated again with the same methodolgy you used this time. Log files have their own sets of problems!

By: walt crawford

walt crawford — Mon, 15 Aug 2005 15:08:20 +0000

Rochelle, TangognaT:
“Really, the best way to do a survey like this is to ask the people involved to give up their web stats.”

You may be right, but that’s not going to happen, at least not for a repeat if I do it:

1. Quite a few bloggers don’t have direct access to their log files (e.g., those using free hosted services, as far as I know–for that matter, I don’t know how I’d deal with my own log files).

2. I don’t have the horsepower to do analysis of, say, 240 sets of log files. I think that one step would double the effort required to do the whole survey, which would doom it–it just wouldn’t be worth the trouble.

3. Log files do *not* reflect RSS readership accurately, as one (say) Bloglines fetch stands in for all of the readers–and it gets worse, because some (not all) of those readers come back to the blog itself. Unless you’re saying that every RSS reader actually leaves behind the number of people it was feeding for.

In any case #2 is the killer: If the general consensus of the community is that a survey is worthless unless it’s based on self-reporting and analysis of all the log files, then I, at least, wouldn’t even attempt such a survey: The cost in time would simply be too high.

By: tangognat

tangognat — Mon, 15 Aug 2005 14:14:38 +0000

Giving up log files might work, but in my case I have so much referrer spam, it makes looking at my stats almost useless.

By: Rochelle

Rochelle — Mon, 15 Aug 2005 13:30:38 +0000

For those who consistently skim their statstics, all of that information is available. Livejournal bots leave the details of how many subscribers they’re collecting your feed for, and you can see the hits from individual feed readers. Really, the best way to do a survey like this is to ask the people involved to give up their web stats.

By: Stefan Hayden

Stefan Hayden — Sat, 13 Aug 2005 18:34:58 +0000

Right now all us techy people seem to be more focused on getting people to use RSS then counting. I was always surprised that such stats were not more part of early rss readers. I think as time goes on more aggregators and blog engines will focus more on RSS statistics so that all the different ways to publish and read will play nice together.