Tracking RSS subscribers via the IMG tag – a quick Web Analytics HACK

The single most intriguing KPI running a blog is: “# RSS Subscribers” – and most of us, get our daily fix through FeedBurner or a similar service. BUT there is typically a disconnect between this essential blog KPI and the remaining KPI’s, which usually are tracked by ones preferred Analytics Tool such as IndexTools, WebTrends or Google Analytics.

A typical reporting on the # RSS Subscribers – KPI looks like this:

note: RSS subscribers trend from 21st June 2007 to 4th July 2007 from the feedburner report.

However; there is very little opportunity to do analysis on this KPI, given the environment where we get it from — and the limited data that it can be paired with — thus we end up looking at the reporting only. Therefore, it is important to move this KPI into ones standard web analytics tool and there is a quick and dirty Web Analytics HACK to do so :-)

Set up an Action (Goal) for you to track the # RSS subscribers. This is how it is presented under your IndexTools action settings:


Customize the noscript pixel part of your tracking code – so that the previous created custom action is recorded. The syntax is as follows in e.g. IndexTools:

a = the project that you want the collected data to be stored in
x = the action you want to set


Here is where you can locate this IMG tag in the code. this is quite similar from tool to tool.

In RSS 2.0 you would include the above URL as an image as part of your channel definition or in ATOM you would include the above URL as part of your atom:logo definition. If you like me, run a service like FeedBurner, there is no need to fiddle around with the above RSS or ATOM syntax, you simply specify a “Feed Image” for your FeedBurner Feed. This is where to paste the above URL in FeedBurner:

That is it! – you now have a “real-time” # RSS subscribers metric available in your Analytics tool.

NOW! the bigger question; Without even describing how to actually define and conclude on what a RSS subscriber is – I take for granted that the delivered subscriber number from FeedBurner is somehow a de-facto standard and trusted. That said; comparing the number of subscribers as delivered from FeedBurner with that collected through the IMG tag; we get the following result (for this blog):

Is this accurate data; NO! – does it matter; NO! :-) – what matters, as most of the industry have been preaching for years, is the trend – as long as you are doing analysis! It should be said that if you are generating revenue of your RSS subscribers you would need the most accurate top line number of course. Just by looking at the above FeedBurner vs. IMG tag comparison, one can see that the Trend is very similar. Yes! there is a factor 4 difference in the actual numbers, but why would we care about that in most of the analysis that we are doing?

By adding a linear trending on the dataset we will get and even better idea on whether we can use the IMG tag for any serious analysis.


This shows (or actually my RAW calculated trend line numbers) that the GROWTH percentage is 26% when looking at the IMG tag data and 15% when looking at the FeedBurner data. However; looking at the data, any linear trend line will be a tad off with the limited amount of data and this much fluctuation. I am actually quite confident that this will become even more accurate from a trending point of view, when I get more data, as a lot of the fluctuation can be sorted out by looking at it from an e.g. week basis.

So the conclusion is — and this is based on a 14 day short data sample, and I might change my mind — That I believe it is very reasonable to use this quick hack to start doing some RSS subscriber analysis! – As in e.g. where are my readers located?


Paris? – I am actually wiser now! :-) …. Bonjour, je m’appelle Dennis.

Another great article on the Subject (Webtrends example included) is:
Web Analytics and Feeds #1: Feedburner (by: Eric Butler)

  • Anonymous

    I guess this will work fine with IP+Useragent tracking. I wonder if this hack will work with cookie-based tracking… for example, Google Analytics.

  • EricB

    Well said Dennis…thanks for the post, and for the kudos. Cheers.


  • Dennis R. Mortensen

    Hi there/..

    >>I guess this will work fine with IP+Useragent tracking. I wonder if this hack will work with cookie-based tracking… for example, Google Analytics.

    I am assuming that most of us have moved on forward and are using some sort of page-tagging (or cookie-based tracking as you quite correctly call it) – This including Google Analytics. Most of these Web Analytics solutions; whether that is IndexTools, Omniture, Webtrends or even Google Analytics (as far as I remember) – have a HTTP user-agent IMG request method in place (partly as a “fall-back” methodology as well). AND it is this IMG that can be used in this “HACK” :-)

    For Google Analytics, the tracking pixel is called “__UTM.gif” – thus a call in this direction: “…”

    Please correct me if this is outdated. However; I am quite sure it is correct. Please post the syntax when you created it.


  • Dennis R. Mortensen

    Hi Eric

    I would love to hear how your collected data compares to your FeedBurner data.. since I only have 14 days worth of data and it seems like you might have a year.. :-)


  • EricB

    I’ll have to look into that more closely when I get a chance!



  • Maria Susu

    the solution is not right .. google analytics moved on to a diiferet paradigm – the img is no longer used

  • Dennis R. Mortensen

    Hi Maria,

    hmm. I am not so sure you are right. I (.. and remember I work for Y! not G) still believe that the UTM.GIF is still called and that the old syntax such as the below is still valid
    utmdt= &
    utmcc=__utma%3D<__utma cookie>%3B%2B__utmz%3D<__umtz cookie>%3B

    ..or did I completely misunderstand you :-)
    Cheers and thank you very much for commenting. Please let me know if I am not right here. (because it is actually simple to check)