RSS Subscriber Analyzer

about | archive


[ 2005-March-11 14:11 ]

This is a simple Python script that attempts to determine how many people are subscribed to your weblog. It is definitely not 100% accurate. You feed it lines from your web server log, filtered to only show hits for your RSS feed. It reports the number of times your feed was downloaded by different users. The algorithm is very simple. If the hit is from a service like Bloglines or NewsGatorOnline, the number of subscribers they report is added to the daily total. Otherwise, we count unique (ip address, user agent) pairs. This means that the number will be underestimated if two users with the same program access your feed from the same IP. The number will be overestimated if a user accesses your feed twice in the same day from different IP addresses.

Download: rss-analyze.py

Usage is something like: grep 'index.rss' /var/log/thttpd.log | python rss-analyze.py

Output:

1/Mar/2005	62
02/Mar/2005	72
03/Mar/2005	78
04/Mar/2005	78
05/Mar/2005	69
06/Mar/2005	68
07/Mar/2005	84
08/Mar/2005	80
09/Mar/2005	85
10/Mar/2005	69
11/Mar/2005	70

Wow! I have somewhere between 60 and 80 subscribers. Thanks for reading everyone.