Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
This forum is in permanent archive mode. Our new active community can be found here.
Feedburner Stats
I know you guys like your feedburner stats. However, how accurate do you think they are? From my own experience, when we moved our podcast from our own hosting and feedburner, to Libsyn, our average audience shot up by around 300 people per week. As we changed nothing with the show, or our website, I'm assuming that feedburner actually records low. Or is it that Libsyn records too many listeners?
Comments
But as you get more and more traffic the law of averages starts to take over. You can begin to look at not the stats themselves, but the relative numbers. If you had 400 unique ips download the file yesterday and 500 today you probably did not gain 100 listeners. But you gained a lot for sure. Just check to make sure there wasn't some technical error or such that caused 100 extra somewhere. Also check for referrals to see if a popular site linked to you.
As for the difference between libsyn and feedburner it's rather easy to explain. Feedburner just has your feeds. So it can only count listeners who subscribe to your show with podcatchers or listeners who go to feedburner in their browsers. Libsyn hosts your mp3, so they can count actual downloads. The difference between the two are the people who are listening to your show without podcatchers.
It's not statistically reliable to really compare two completely separate sources of data. Nor are any of the raw numbers reliable to any degree. Web statistics can only give you a general idea of gaining or losing listeners. Don't trust the real numbers without a lot of data and a statistician.
I don't get Scott's dig into me, though ;^) I see them once a day because I have to log in to the site in order to upload the file anyway. (I'm not using cleartext ftp over the Internet).
As for trends in data, they tend to emerge rather quickly. After about two weeks worth of podcasting way back, the data stabilized and have been largely predictable since.
I observe them regularly to ensure that the trend is always upward. Downward movement over any reasonable span means something's wrong. Happily, the plotted graph of our listenership is an almost perfect increasing linear correlation.
I've had great results tracking refers with a free stats program TraceWatch (www.tracewatch.com). At the moment on our site it's only tracking visitors who have javascript enabled on their browser, but it can track php pages with a big of code. I'm sure you've probably come across this sort of thing before. It will even show the routes that people take through your site and stuff like that.
We do just use Libsyn stats for our podcast though, its so good that it's hardly worth going to the trouble to do it another way. That said, we'll see how they hold up now we're talking to sponsors!