How are podcast downloads tracked and tallied?

As you may know, FeedBurner does not actually host your podcast or video cast files. We do, however, create <enclosure> tags in your feed whose links track podcast downloads (but only if you have enabled this stats tracking option). When we receive a request from someone for one of the enclosure links, we:

1) Count that request as a "download" request

2) Redirect the request to your original media file's location.

Since we began offering download tracking, we have learned much about many "user-agents," or applications, that people use to consume podcasts. The key lesson: they do not all behave the same way!

Some user-agents follow the redirect to download the file and that's it -- end of story. However, some are more sophisticated and download different pieces of the file associated with the URL by re-requesting the URL many times. Each one of these requests is legitimate in that we redirect it to the server hosting your podcast or video cast file and that server has to respond to the request. However, counting these requests does not really help you understand how many individuals are downloading your podcasts or video casts; the numbers become noise.

So here's what we do.

If the same IP address...
requests the same enclosure URL...
from your feed...
using the same user-agent...
more than once in the same 24 hour period...

...we will only report one request.

We only count one request a day per file from a single individual downloader.

This download counting procedure has been in effect since October 4, 2006; podcasters with statistics that pre-date this date have not had prior stats adjusted to reflect this tracking approach.