Saw a mention of AideRSS go by on Twitter the other day, and just tried it out. The idea is great – give
AideRSS a blog URL and it purports to sift through all the posts and then offer separate feeds to you so you can subscribe to only the “best post” the “great”, “good” or all the posts (the naked feed, if you will). Supposedly it’s a way to deal with overload, by only getting the best of the posts from the bloggers you want to follow. Sounds great, right?
Maybe, but IMHO they’ve got some work to do on their algorithms. With someting like this I like to compare myself to someone I know is a “better” blogger. In this case I picked Meredith at Information Wants to be Free. Yeah, we blog differently, but AideRSS appears to place a lot of weight on how many links and how much of a conversation a given blog post has. In that case Meredith definitely blows me away. And yet, here are the screenshots of her blog and mine. Take a look at the highlighted sections. Any rhyme or reason there? Look how many posts, saves and comments her posts have, and she doesn’t even get a “best posts” category! Maybe they’re placing way more weight on # of recent posts than they should? (click for bigger versions)
By all means go give it a try – maybe it’ll be useful to you.
Comments
2 Responses to “AideRSS – Great idea, but…”
Hi DL — Thanks for giving us a test drive!
Right, so, a few issues to address (will try to keep things as brief as possible).
So… Basic analysis runs on what we call our Feed-based PostRank algorithm. It’s purely apples to apples, i.e. looks at your own site’s performance over time (or whichever site you’ve plugged in to analyze). So those rankings aren’t being compared with any other site.
We do also have Thematic PostRank analysis, which is more apples to oranges. Currently that appears in our Google Reader extension when you’re in folder view, as it compares the performance of all the feeds in the folder against each other. We’ll be expanding the availability of Thematic PostRank over time.
So really, comparing your stats to Meredith’s doesn’t really mean anything. For example, if posts of yours usually get 2 comments each and then one post suddenly gets 12 comments, you’re going to see a big PostRank jump for that post. However, if another blogger gets 12 comments for a post, but most of her posts usually get 100 comments, she’s going to see a big PostRank drop for that post.
The missing Best Posts section for Meredith’s blog is… not really a bug, but not an optimal process. Basically, those percentages look at how many posts from a blog that we’ve analyzed meet the filtering parameters for Good/Great/Best/etc. If less than 1% of a blog’s posts are above a filtering threshold, then 0% appears, as in your screenshot above. I certainly don’t like how that works since it’s really not accurate.
Fortunately, the issue will be moot shortly, since we’re planning on releasing the next version of our site next week, and there’ve been a number of changes, including dynamic generation of filtering, so an issue like “Best = 0%” will no longer occur.
Hope that makes sense. If you have any questions, feel free to give me a holler.
Thanks for the explanation Melanie. I look forward to the new release, and will give it another look then. 🙂