10 Comments

  • Andre Ward

    I also read about this on ProBlogger, I love the service. Proves a handy way of analyzing your works and other peoples works.

    Truly something people should consider using often.

  • Really cool! I gonna give it a shot.

  • As you said in the post, the problem is that it’s missing many thing, StumbleUpon among them. For instance, once I got thousands visitors for a post that only got a score of 7 out of 10 using their service.

  • @Andre Ward

    Yep, what I like in concept – it does something that it’s pain to do manually. Computers are good at doing such boring things like gathering and comparing data.

    @Ky

    Problem with number of visitors is that only website owner has reliable access to that data. :)

    AideRSS clearly far from perfect in current state but I like core idea and if they keep working at it (which they seem to do looking at activity on support page http://getsatisfaction.com/aiderss/ ) it may mature in truly great tool.

  • I’m not saying it’s bad, just that AideRSS could be greatly improved.

  • @Ky

    Yep, that’s exactly my opinion as well. :)

  • Thanks for the very comprehensive write-up!

    The issue with missing comments and such is a pain, but unfortunately since there’s not just one reason why that can happen, we need to address it on an individual feed basis, so we encourage everyone to ping us either on the Get Satisfaction site or email me.

    We do keep adding new sources of analysis, but unfortunately we are limited by the data availability. We’ve received a number of requests for StumbleUpoon, for example, and would love to add it, but have yet to get access to an API.

    For those who can tweak their RSS settings, we can display a great many items (more than just the 10 or 50 or 100 that seem to be standard).

    Re. the rankings, the popularity of a single post isn’t quite the whole story. For example, if one post gets a flood of bookmarks or comments or what have you, yes, it will certainly come up scoring high, but at the same time we analyze it in terms of past site performance, so we take anomalies like the “digg effect” into account so that it doesn’t completely skew all the other posts’ results (automatically making them 1s, for example). And if there’ve been other posts that have done well recently, then compared to them, a highly engaging post will do very well, but won’t be head and shoulders better than the previous ones. Hope that makes sense.

    As always, we’re always interested in questions and feedback as we continue to work on improving things. We will also be launching the next version of our site soon, which will have a lot of really cool new and improved functionality. Stay tuned!

  • @Melanie Baker

    Thanks for dropping by. :)

    If I may voice my opinion about data issues – service is not clear enough on data sources and how they influence score.

    For example I got 10 on this post
    https://www.rarst.net/web/engineering-windows-7-ray-of-real-news-on-os-future/
    but there are only two comments on it and all other values shown are at zero. I understand that something had worked with this post but I have no data on what it was.

  • Well, unfortunately the explanation very quickly gets into the propriety stuff, so it’s never been easy to offer one that people are totally happy with. :)

    One “theme” re. how rankings work is that PostRank scores are calculated based on a number of posts, not based on a span of time. Another is that within a group of posts being analyzed, with relatively consistent engagement performance, PostRanks are most inclined to rise, rather than go up and down like a typical graph might.

    At the moment we’re also working on fixing delicious analysis as well. Some things stopped working when they released their new version.

  • Still to please power users you’ll have to offer some details. :) Probaly not like google analytics or feedburner (overkill) but some…

    I’ll be waiting for that next version, good luck developing. :)

Comments are closed.