RSS Feed Server Stat Hits

January 12, 2006

I noticed a surprising trend on one of the larger sites I currently develop and maintain. I check the server stats daily and use three different methods: Webstats (provided by the hosting company), Urchin stats (provided by the hosting company), and Google Urchin (enabled via javascript). While we consider pageviews to be the most important indicator of traffic, by default I also check the hits and requested files stats.

What's happening is that our RSS feeds are starting to register as some of the most widely hit "pages" on the website. We have four different RSS feeds on the site, three unique RSS feeds, and a main RSS feed which incorporates the top entries from the other three (duplicate content, I know). Right now, the four RSS feeds are all in the top 10 most requested pages on the website (does not include images, style sheets, etc.), and rank very high in terms of files by hits (includes images, style sheets, etc.).

I doubt at the moment many people subscribe to the RSS feeds. The site came into existence this past fall (about 5 months ago), and was completely redesigned by myself and went live the day before Christmas last year (December 24, 2005). The "old" site did have an RSS feed, but it was only one, and it's location was in a different directory than the four we have now.

Further inspection of the stats show that Feedreader, an RSS reader that no one in our office uses, is the second most popular "bot" hitting our site, and hits our site more often than all the search engine bots (including MSNbot). Two people in our office (including myself) use RSS Reader which registers in about 1/10 the hits Feedreader does. Just below RSS Reader is Feedfetcher Google, the Google RSS reader.

My question is, do these bots and RSS readers count as actual server stats, server hits, and in essence, server traffic? With the basic denomination of making money from a website deriving from server traffic, doesn't it seem easy to "artificially" increase traffic through the use of RSS readers? Furthermore, most RSS readers allow you to set the interval at which they check for updates to the RSS feeds they check. So if you increase the duration of checks, will that increase the server traffic? If all of this true, shouldn't it be very easy to create traffic to a website by having a number of associates subscribe to your RSS feed and set the interval to check updates as low as it can go?

As our site gains popularity and more people subscribe to the RSS feeds it will be interesting to see if any of the above holds true. In an industry where server traffic and page views dictate how much you can charge for advertising (or if indeed anyone will advertise at all), it seems like a site with a large RSS subscriber base would far exceed a site with no RSS feed offerings in terms of traffic.

The next question becomes, should RSS reader hits count towards server traffic? I think the answer is yes, and no. Certainly when my RSS reader checks for new entries and presents me with one, I'm apt to read it, and I think that should count as a hit. But when an RSS feed is dormant for a week, I do not think the updating hits should count for traffic.

I don't think many advertisers are privy to this data, and when they simply ask for a monthly page views count, I suspect very few if any ask how many are attributed to RSS readers. The lesson learned is that if you are an advertiser and are considering a site to advertise on, you would be wise to get the raw server stat logs before making any decisions.

Technorati Tags: rss feeds  server stats  server statistics  server statistics analyzing