ADR - Average Daily Readership
One of the RSS Metrics that is extremely sought after by RSS Feed publishers is the ADR or Average Daily Readership. The method for calculating this value is a bone of contention. I always tell my clients that metrics have to be of value to you. Tell me how you want to define each metric you want collected, and I'll know what to do.
The essential idea behind ADR is to calculate how many unique readers use your feed each day, on average, calculated over a duration. Immediately you wonder, what duration? Well, if your feed has only been available under a month, you will not have an accurate average. It's seems like a good idea to me to recalculate the ADR for a feed at the end of each month. That's certainly as good a place as any to start. But there are other complications that you must decide how to handle.
For example, a number of RSS Readers have a feature whereby you can automatically update all your subscribed feeds on a defined frequency, say every half hour, or hourly, etc. If you have subscribed to, say, 20 feeds, and you update every fours in a given work day, at work, that's probably 2-3 updates each day. Assuming the IP address of your computer doesn't change, each publisher's web server logs will record the fact that your IP address has accessed their RSS feed 2-3 times each day. However, that by no means says you have read any of the items in their feed.
A more accurate indicator is to match each unique IP address in the server logs that has accessed the main feed's URL, plus has accessed one or more of the "details" pages. Say a feed has 15 current items, and you update your feeds twice daily. But you do not have time to read any items until the afternoon on most days. Your RSS Reader is such that the first item of the selected feed automatically gets "read". (What I mean by this is that most standalone RSS Readers bold the title text of feed items that are unread, then unbold the text if the item gets selected either by a mouse click or using the arrow keys.) In a standalone reader, this "reading" of the item DOES NOT send a request to the publisher's web server. Instead, standalone readers have already downloaded a copy of each feed you subscribe to. But when you actually do read, say, 5 items during your afternoon break, you might click through to, say, 3 items. Each clickthrough of a feed item now generates a request to the publisher's web server.
On each clickthrough, you will be served up the "details" page of the item. Since you clicked through on three items, that's 3 visits to the publisher's web site. But you are only one person, so for that day, you should only be considered once the ADR metric. However, in terms of RSS visits for that day, you should be considered 3 times for the RSS visits metric. Understand?
In the next posting, we'll see why the claim that publishing an RSS feed increases web site visits is a bit misleading. And I still promise to talk about how to measure readership when your feed is being syndicated by some other web site. In other words, if some of your feed subscribers get your content via a web-based RSS reader service instead of a standalone application.
The essential idea behind ADR is to calculate how many unique readers use your feed each day, on average, calculated over a duration. Immediately you wonder, what duration? Well, if your feed has only been available under a month, you will not have an accurate average. It's seems like a good idea to me to recalculate the ADR for a feed at the end of each month. That's certainly as good a place as any to start. But there are other complications that you must decide how to handle.
For example, a number of RSS Readers have a feature whereby you can automatically update all your subscribed feeds on a defined frequency, say every half hour, or hourly, etc. If you have subscribed to, say, 20 feeds, and you update every fours in a given work day, at work, that's probably 2-3 updates each day. Assuming the IP address of your computer doesn't change, each publisher's web server logs will record the fact that your IP address has accessed their RSS feed 2-3 times each day. However, that by no means says you have read any of the items in their feed.
A more accurate indicator is to match each unique IP address in the server logs that has accessed the main feed's URL, plus has accessed one or more of the "details" pages. Say a feed has 15 current items, and you update your feeds twice daily. But you do not have time to read any items until the afternoon on most days. Your RSS Reader is such that the first item of the selected feed automatically gets "read". (What I mean by this is that most standalone RSS Readers bold the title text of feed items that are unread, then unbold the text if the item gets selected either by a mouse click or using the arrow keys.) In a standalone reader, this "reading" of the item DOES NOT send a request to the publisher's web server. Instead, standalone readers have already downloaded a copy of each feed you subscribe to. But when you actually do read, say, 5 items during your afternoon break, you might click through to, say, 3 items. Each clickthrough of a feed item now generates a request to the publisher's web server.
On each clickthrough, you will be served up the "details" page of the item. Since you clicked through on three items, that's 3 visits to the publisher's web site. But you are only one person, so for that day, you should only be considered once the ADR metric. However, in terms of RSS visits for that day, you should be considered 3 times for the RSS visits metric. Understand?
In the next posting, we'll see why the claim that publishing an RSS feed increases web site visits is a bit misleading. And I still promise to talk about how to measure readership when your feed is being syndicated by some other web site. In other words, if some of your feed subscribers get your content via a web-based RSS reader service instead of a standalone application.
| Permalink | 
