is it possible to measure rss subscriptions?

Blog stats aren't, of course, everything. Although I use statcounter on cybersoc.com, I'm a lot more interested in seeing where people came from, and where they go once they're here, than I am in knowing the shere numbers of visitors. That said, the stats for unique users, returning users and page views are there and I'd be lying if I said I didn't look at them.

Yesterday, according to statcounter, 244 unique users clicked through to cybersoc.com, generating just over 300 page views. About 15% of those were returning visitors that is, they were people that statcounter recognised as having visited before.

Feedburner, which I use to provide an RSS feed of this blog, says I had 98 subscribers yesterday. I have no idea how they get this count, or if it is a true reflection of the number of subscribers or not. Feedburner breaks those subscribers down by the RSS readers they use (see left).

According to feedburner's breakdown of subscribers to my RSS feed, there were 24 bloglines users subscribed yesterday. I happen to be a bloglines user myself, so I pretended I wanted to add my feed to the feeds I subscribe to. For those not familiar with bloglines, you simply type the URL of the blog or page you want to subscribe to and it goes off and gets you a list of available feeds and tells you how many people are subscribed to each of them. Bloglines reckons I have over 100 subscribers - and this just lists the people who are subscribed via Bloglines.

So if bloglines users equate to 25% of the subscribers to my RSS feed, and there are over a hundred subscribers on bloglines, then shouldn't the number of subscribers to my feed equal something more like 400?

I've long had an inkling that the number of people reading this blog via RSS is similar or greater, because of it's likely audience, to the number who actually click through and read the "ordinary web page" version. I base this on the number of comments and emails, as well as links on other blogs, that I get from people who tell me they subscribe to the RSS feed.

I'm interested to know if anyone has learned how to better track the number of people who read via RSS readers. I'm not terribly concerned about it for the purposes of cybersoc.com but, rather, wonder how websites that publish feeds, particularly commercial websites, are going to be able to analyse and report their levels traffic in future.

If unique users and page impressions remain the standard measurements of the industry, than sites that don't offer RSS feeds, or that have a less technologically savvy audience, may actually be rewarded. That said, by not offering feeds, their content could just sit there in a self made content ghetto, unshared and unloved. Or something like that.

Any thoughts on measuring RSS subscriptions?

Comments

[Ian Delaney]

I believe the Feedburner numbers are 'active' subscribers - that is, people who have loaded your feed today. That's why the numbers go up and down so dramatically. This seems a reasonable metric - after all, a million subscribers who never read your stuff don't really count.

[John Thompson]

Robin

Short answer, yes it is. We are currently testing a service that does just that - http://www.mediafed.com/

A few bugs to sort out (mostly ours) but it has the potential to give us some very useful information about what items get the most views and clicks, plus country of origin of visitors and even what RSS readers are being used.

[RobinH]

Hi John. I've spotted a few of those mediafed links from bloglines to articles on journalism.co.uk. The links seem really slow and it takes a while for pages to load. Long enough that I often think the page has timed out. Might be worth looking into. The article I was trying to see just a moment ago that reminded me of this was here:

http://www.journalism.co.uk/news/story3074.shtml which was redirected first from a feedburner url, then to http://rss1.mediafed.com/feed/journalism/News/?link=227e28a6378f2e175b9b71d0969b9ae6

Hope that helps.

[brad]

Currently the only way I can figure out RSS subscribers is by requested pages.

I'm using Urchin for traffic reporting. It lists RSS feeds in Requested Pages. I take the total for the month and then divide it by the number of posts in the month.

I'm not particularly confident about the numbers, although I did find out that RSS has a kind of 'check for new items' function, which returns a yes or no response. In other words, I'm fairly sure that though my RSS feed reader checks for new items every half hour, none of these get counted as a request for a page. A successful request will only occur if there is a new post.

We could certainly use better ways of measuring RSS. On the site I'm trying to measure, it seems to consistently account for about 20% of all page requests, which is double the second highest ranking page. (Mind you, the 2nd highest page is the result of comment-spambots :-)

It seems the more I look at stats, the less confident I am that I know anything.