How many people you think they got working on that Feedburner thing over there? One? Half? A summer intern? A guy who works on it while he watches football?
The flaws and problems with Feedburner have been frustrating to say the least, so I was pretty damn excited when, a little over a month ago, Feedburner finally announced that they were integrating the ability to add in Google Analytics campaign data to the end of feed item URLs. This was huge. Before this, you had to add in the string into the original RSS feed, which most of us couldn't do since we didn't have access to that original feed, but wanted to track when people were coming from our feeds.
My excitement ended, though, when I cranked up the feature and saw what it did to my stats. Yes, it was helpful to see who came from Google Reader or Bloglines, but if the content was syndicated, it stripped out the referrer. This was a problem with my music blog. Hype Machine is by far and away our #1 referrer, and when I implemented this change, the clicks from there went to almost none, because the campaign source was hard-coded in there as "feedburner". Which was not helpful at all. If the campaign is already "Feed: {etc}", then what does having the source as "feedburner" tell you? Nothing, that's what.
My Solution
When the campaign feature was first announced, I noticed an interesting dynamic code in the "content" area: ${distributionEndpoint}. This made much more sense in the source area, so I put it there instead of "feedburner", only to find that the only distribution endpoints that Feedburner acknowledged were Google Reader, Google Feedfetcher and Bloglines. Now, wouldn't it just make sense to have this include whatever site referred it? Yes. But that's Feedburner for you.
But what happens when it doesn't acknowledge one of those few endpoints is that it just leaves it blank. So what I did, see, is to go ahead and put ${distributionEndpoint} in the source, so that when the feed doesn't find Hype Machine (which it probably won't for a while) or any of the other endpoints that it won't find, it just uses the referring site as the source like normal. And my referrals from Hype Machine are back to where they were.
By the way, I hate to badmouth Feedburner so much. It really is a really good service. But I get really frustrated with some of the half-assed implementation that Feedburner (among other Google properties) implements, suggesting that the services are not really much of a priority for Google.
The flaws and problems with Feedburner have been frustrating to say the least, so I was pretty damn excited when, a little over a month ago, Feedburner finally announced that they were integrating the ability to add in Google Analytics campaign data to the end of feed item URLs. This was huge. Before this, you had to add in the string into the original RSS feed, which most of us couldn't do since we didn't have access to that original feed, but wanted to track when people were coming from our feeds.
My excitement ended, though, when I cranked up the feature and saw what it did to my stats. Yes, it was helpful to see who came from Google Reader or Bloglines, but if the content was syndicated, it stripped out the referrer. This was a problem with my music blog. Hype Machine is by far and away our #1 referrer, and when I implemented this change, the clicks from there went to almost none, because the campaign source was hard-coded in there as "feedburner". Which was not helpful at all. If the campaign is already "Feed: {etc}", then what does having the source as "feedburner" tell you? Nothing, that's what.
My Solution
When the campaign feature was first announced, I noticed an interesting dynamic code in the "content" area: ${distributionEndpoint}. This made much more sense in the source area, so I put it there instead of "feedburner", only to find that the only distribution endpoints that Feedburner acknowledged were Google Reader, Google Feedfetcher and Bloglines. Now, wouldn't it just make sense to have this include whatever site referred it? Yes. But that's Feedburner for you.
But what happens when it doesn't acknowledge one of those few endpoints is that it just leaves it blank. So what I did, see, is to go ahead and put ${distributionEndpoint} in the source, so that when the feed doesn't find Hype Machine (which it probably won't for a while) or any of the other endpoints that it won't find, it just uses the referring site as the source like normal. And my referrals from Hype Machine are back to where they were.
By the way, I hate to badmouth Feedburner so much. It really is a really good service. But I get really frustrated with some of the half-assed implementation that Feedburner (among other Google properties) implements, suggesting that the services are not really much of a priority for Google.
Comments
Post a Comment