Post Rank uses a variety of ways to measure levels of “engagement” that readers have with specific blog posts. Recently, they’ve begun a system to tabulate this data and develop lists of the most popular blogs in many different categories. The rankings are adjusted each week or so.
I thought it would be useful — both to me and readers of this blog — to do a quick review of the top five blogs in several categories that might be useful to those of us in the education field.
I have the phrase “most popular” in quotation marks because, of course, Post Rank’s system is just one of several ways that you can gauge “popularity.” I’d just use this list as an opportunity to explore some new blogs.
Sue Waters, the person whom I most respect in the education blogosphere, raises a “red flag” about Post Rank:
“My argument still remains that PostRank’s system of ranking importance of most popular posts is flawed so would hate to think how they are determining what’s the most popular blogs in each category.
The trouble with all these systems is that people see the words “Most popular” and don’t stop to consider the statistics and methods used to provide the ranking. If the underlying system is flawed then the list is.
What is the solution? Not sure. If technorati was working properly than its a better reflection as is number of subscribers using Feedburner. Until someone gets PostRank to properly explain properly how they are determining what’s the most popular blogs in each category I believe we should be viewing their data with extreme caution. And while it is okay to use their data I would be emphasizing strongly this aspect.”
Given those caveats, here are (in my view — in that I’ve chosen the categories –and in Post Rank’s view since its their rankings) The “Most Popular” Blogs That Might Also Useful To Educators (the heading of each section is a live link that will take you to the complete listing and ranking of blogs in that category) They’re also in the order that I thought of them or discovered them — which means the categories aren’t in any order at all:
1. Free Technology For Teachers
2. Live Science
3. This blog, Larry Ferlazzo’s Websites Of The Day…
5. Strange Maps
2. Pharyngula
4. Discover
5. Epicenter
3. Techie Buzz
4. Mashable
1. Web Urbanist
5. Yanko Design
3. New York Times — Technology
3. The Oil Drum
4. paid Content
1. Engadget
2. Gizmodo
5. Gadget Lab
1. Geek Dad
3. Play Library
1. Seth’s Blog
5. Sanders Says
5. McSweeney’s
2. BEYOND Bones
5. Egypt at the Manchester Museum
2. Flowing Data
4. Fubiz.net
5. You the Designer – Graphic Design Blog
1. PSDTuts – Just Great Photoshop Tutorials
2. Instructables: exploring, featured
3. Dumb Little Man – Tips for Life
5. Photojojo
Suggestions and feedback, as always, are welcome.
If you found this post useful, you might want to look at the 250 “The Best…” lists (including The Best Places To Find Good Education Blogs) and also consider subscribing to this blog for free.
My argument still remains that PostRank’s system of ranking importance of most popular posts is flawed so would hate to think how they are determining what’s the most popular blogs in each category.
The trouble with all these systems is that people see the words “Most popular” and don’t stop to consider the statistics and methods used to provide the ranking. If the underlying system is flawed then the list is.
What is the solution? Not sure. If technorati was working properly than its a better reflection as is number of subscribers using Feedburner.
Until someone gets PostRank to properly explain properly how they are determining what’s the most popular blogs in each category I believe we should be viewing their data with extreme caution. And while it is okay to use their data I would be emphasizing strongly this aspect.
Larry,
I’m with you and Sue, I’d love to know how Post Rank really calculates their rankings. While it’s nice to see myself at the top of the list, it doesn’t really mean much unless I know what the basis for that ranking is.
Technorati is a good concept, but the functionality of Technorati has been lagging for some time. The last time I checked Technorati it said I hadn’t posted anything to my blog for 15 days. As we know, I have trouble going 15 hours without posting let alone 15 days :).
All of the above being said, I like looking at lists of blogs because you never know what new thing you might discover. In that regard, any system that produces ranked lists does have some value to me.
Richard
Thank you, I’m flattered to be on this list.
Daniel Tunkelang, The Noisy Channel
http://thenoisychannel.com/
You and your lists, Larry! They are awesome!!! Thanks for putting this together. 🙂
Hey, great list and blog Larry!
What about the site http://www.kidsstoriesonline.com/ for your list
While it sure feels good to see my blog so high up on a list like this I also wonder about how the ranking is calculated. Several of the other education related blogs they list below mine have far larger followings and far more links as shown in Technorati. Still I do like the conversations that happen at my blog especially lately. And if a list like this sends a few more readers my way I will not complain. I make it a point not to complain when I don’t make Top x” lists either. They all serve a purpose. They are all somewhat arbitrary. And they are all subject to change. For me I just try to do my best to be helpful and informative.
Here is a bit more food for thought.
If you go to a blog’s post and closely analyze each statistics there is an eye as one statistic. That eye represents the number of users that have clicked on their postrank widget on your blog.
Now go through and analyse the top blogs. If you examine closely you will notice most of them rank highly based on having high number of clicks on the links in Postranks widget.
They also will rank posts higher when it has lots tweets. For example I’ve seen posts get 10 for large number of tweets without it necessarily being a great post it is just we want to help someone with their child’s project.
If you think about it really carefully, while PostRank provides a helpful resource for quickly working out number of Delicious saves, comments, and links (and not that accurately), its actual system of ranking posts and comparing blogs ranking is flawed.
So the answer is if you want to rank highly with PostRank you add their widget to your sidebar. mmmm not good!
Sue,
Thank you for the explanation. That makes sense to me. I’ve often thought that it had something to do with the number of times a link was shared. That explains why “list posts” are often the most highly ranked posts according to Post Rank.
Richard
Hi Richard, I’ve also followed up with a more detail post on my personal blog. The true power of PostRank is being able to analyse the performance of posts on your own blog.
Have a look at the screenshot at the bottom of my post and you can see why using it as a comparison between blogs and each bloggers post is worrying.
It’s great to see so much interest in PostRank, though there are a few misconceptions flying around, so let me see if I can help with that. 🙂 (This will be quite long for a comment…)
It certainly looks like you folks have the same discomfort with the word and implications of “popularity” as I do. And so we tend to avoid that word, with good reason. No one knows what it’s supposed to mean, and it’s completely subjective. And really, there’s no room for subjectivity if you want your system to have credibility. We want our analysis to be about algorithms, not opinions.
Unfortunately, “engagement” is also a word that’s getting overused and misinterpreted and is starting to get a bit mushy. In our world, engagement refers to action — how interesting/inspiring/enraging/etc. did people find your content and to what degree did they express that with actions? Writing their own blog posts in response, leaving comments on your blog, bookmarking, tweeting, etc. We outline how PostRank works here: http://www.postrank.com/postrank.
Additionally, we recognize that not all actions represent the same level of engagement. Simply reading your post doesn’t show the same interest and commitment and amount of work as leaving a well-thought-out comment or writing my own response. That’s why we have the “5Cs” hierarchy.
Things like clicks and views are pretty low engagement, and the ones we track via our widget even less so, since not all blogs in our system use the widget, so those are tracked, but not weighted very high. (So Sue need not be so concerned about them being the primary measure of engagement or popularity. Having a lot of “eyes” will not automatically get you a 10 ranking.)
Another step we take to keep rankings relevant is that we don’t take every post a blogger has ever published into account for rankings. (We’ve done research and have calculations to determine engagement and content relevance curves, but I won’t get into that.)
The simplest illustration for that is that blogs change and evolve over time. When you started your blog, a post might have been doing really well if it got 3 comments (to pick one metric). But a year later, you have a lot more readers and have built more of a community, so a high-performing post might get 30 comments. You wouldn’t want those 2 posts compared on equal footing. So comparing the metrics of 2 posts published 6 months apart, for example, is kind of comparing apples to oranges. (We will, however, have trending and historical reporting functionality in the analytics package we’re building for those who do want to see long-term performance stats.)
The trouble with Technorati, or any ranking source that only uses one source of metrics, is that it shows very little of the overall picture. Back in the day when the online world was rather simpler, and we only had links to track, measuring on links made sense. But now we’ve added the entire social web infrastructure and innumerable apps, and we need to take all that into account.
Same with a service like FeedBurner. It tracks subscribers, and, based on all the complaints I’ve been hearing for months, not very reliably anymore, which is unfortunate, since they offered significant value and metrics. Plus, it only works for those who use it. If you’re a FeedBurner subscriber and I’m not, but we write blogs in the same topic area, I can’t really do a FeedBurner-based competitive comparison between our content performance.
The way we try to tackle the metrics issue is by including as many sources as we can. We’re pushing a couple dozen now, with a backlog of a bunch more to add, and we’re always on the lookout for international sources, since the standard social sites, etc. in North America often don’t have the same penetration and influence overseas. Unfortunately, there are some sources we can’t get right now and would love to have, like StumbleUpon, but at the same time, no one else has that data either.
The other thing is that our topics are only as good as the contents in them. All our system feeds are user-submitted, which is actually a great quality measure, and prevents clogging things up with thousands of splogs and whatnot that you can end up with with automated discovery. Additionally, global topics consist of the feeds in our system that have been tagged by users with those keywords. So the topic lists can always get better.
Fortunately, we’ve bet on our community to help with those improvements, so any user can add feeds to the system (even without an account), and any user can add a feed to a topic (needs and account and being logged in). So if the Education topic, for example, is missing someone you know writes great content, by all means — add them in!
When the latest AdAge Power150 list came out recently, with PostRank added to their calculations (http://adage.com/power150/), the ranking changes really illustrated what we’ve found to be the best formula for high rankings:
– blog regularly/consistently
– blog often
– make it easy to organize, share, and respond to your content
– build your readership and community by engaging with them – respond to their questions, comment on their thoughts, thank them for their interest, etc.
Anyway, I could say a lot more, but I’ve already hijacked Larry’s post quite enough. 🙂 If anyone has any additional questions, feel free to post them here (I’ve got comment notification turned on) or to email me directly: melanie@aiderss.com. Or if they’re short, I manage the @postrank and @aiderss Twitter accounts as well.
Hi Melanie, as I’ve said in both my blog post and email I love PostRank in terms of it’s an incredibly valuable tool for quickly analyzing and comparing all of your blog posts in terms of number of:
1. Comments
2. Bookmarked (Delicious, diigo etc)
3. Twittered
4. Linked to
While the numbers it provides isn’t totally correct it is adequate to work out which posts have achieved your desired outcome.
But I disagree with “Having a lot of “eyes” will not automatically get you a 10 ranking” If you go to Larry’s top posts on PostRank and hover each of his rankings on posts almost all of his 10 ranking posts only have the eyes on them (and that is the same for most of the Edubloggers who are using your widget).
Now the eyes represent the clicking of titles of posts in your PostRank widget on a blog side bar. Naturally people visiting Larry’s blog are going to click on a post on the list — which then pushes it up the ranking because they are now assuming that it is a top post. When in reality it is only a top post based on people clicking on the link in your widget.
I’m sure most bloggers definition would totally disagree that a post that is all clicks on links on your widget is a top post and worthy of a 10. Trouble is they don’t realise that is what is happening because they are just looking at 10 and saying WOW my post is doing well.
I think most edublogger would be outraged (maybe I’m wrong) if they analysed their high ranking posts and realised almost 100 % of their top 10 posts are made up of those lovely eyes.
What they need to be doing it saying great my post is a 10, that 10 is only ranked against posts on my own blog and now lets look at what 10 is made of. How much of that 10 is bookmarking, how many is comments, how many are links and to a lesser extent how many are tweets. Then consider how those value compare to what outcome you wanted.
Then if they want to compare their post with another blogger they need to go that persons ranking and hover over their numbers to see counts.
Metrics like bookmarks, links and comments are all good measures of reader engagement. They tell a blogger they are or aren’t achieving their desired outcome. If those are high then sure you are doing well with reader engagement.
But eye clicks on a PostRank widget show nothing in terms of reader engagement except that your readers are going deeper into blog posts. That metric is good for PostRank and helpful for bloggers who want to ensure their blog is high on your PostRanking system.
This comment comes with a health warning: there is a degree of cynicism in it.
First of all, I would like to thank Larry for his post. I found it very thought provoking, and I think that Larry’s aim is praiseworthy. Secondly, I would like to thank Sue Waters for bringing this post to my attention.
Now for the cynical bit …
My primary interest (as an educator) is to provide my students who are interested in Web 2.0 with the resources that they will find most useful to them. Larry’s post is invaluable to me in this matter: I can use it to help my students select the resources that they wish to use.
Broadening this point slightly, there is a wide range of Web 2.0 technologies already in existence, and this range is growing rapidly as each organisation attempts to out-compete everybody else. (This idea is also known as having the “killer application”, one that is so far ahead of the competition that the competition withers and “dies”. Google as a search engine is an example of this.) Without digging into the methods used to derive the statistics used in establishing “the top 5” (and that applies just as much to blogs as it does to the popularity of certain types of cars), I would presume that the statistics are both unbiased and reasonable. (An unreasonable statistic would be “This airline has never had an accident” without you being informed that it has never flown.) Furthermore, given the rapidly changing scene in Web 2.0, a perfectly valid “top 5” at the start of a semester could be hopelessly out of date by the end of the semester. I cite the meteoric rise of Twitter as a case in point.
In summary, statistics about popularity may be of interest to many, but I need to work with a wider remit if I am to provide my students with the best possible learning experience.
I should have been a bit clearer — having a lot of views could, conceivably, result in a 10-ranked post IF that’s by far the source of the majority (or all) of the post’s engagement. (And has been the norm on the blog for a while.) Really, though, that’s not something I see very often. Even low-traffic/engagement blogs tend to get a mix of metrics. But if a site’s readers tend to show engagement overwhelmingly in only one or two ways, then that’s that site’s “culture” and indicative of what the audience considers their best content.
As a side note, there’s a system issue at the moment that we’re working on fixing, which is why Larry’s blog (and others) are behind in showing new posts and some metrics. I see there are some comments and whatnot there, which will have an effect on his scores when they’re processed.
And certainly, if what you’re reporting is that you, Larry, or any other bloggers do have other metrics that aren’t showing up — bookmarks, diggs, tweets, etc. — please let me know which blogs so we can investigate and get that fixed. That I can do something about; differences of opinion re. what “best” is I can’t. 🙂
Re. Naturally people visiting Larry’s blog are going to click on a post on the list — which then pushes it up the ranking because they are now assuming that it is a top post. When in reality it is only a top post based on people clicking on the link in your widget.
This isn’t quite correct. Posts appear in the widget because they already have the highest engagement. There’s an explanation of how top posts in the widget get there here: http://blog.postrank.com/2009/01/07/postranks-website-vs-widget/
So it’s a bit of a “success breeds success” situation — a post resonated with readers, so they engaged with it, which gave it a high PostRank score, which means it ends up showcased in the widget, which makes people more likely to click on it, read it, and further engage with it. For those who already follow a blog, there’s potentially less value there, since they’re already going to be quite familiar with that blogger’s content. For new readers, however, it is a good first glance overview of the blogger’s content, style, and quality.
I must say I love it that the edublogging folks are such an engaged, active, and analytical community. 🙂
Hi Melanie, I would love to valid reasons why PostRank is even including these eye views in the metrics.
Sure they give an indication of how many readers are following those links, and that is nice, but why include when it impacts the metric to this extent?
I’m not sure what I consider valid reasons are what you’d consider valid reasons. 🙂
Ilya (our founder and CTO) and I were discussing this, too, and we do understand where you’re coming from. Really, it’s a matter of managing/balancing edge cases as best we can.
We don’t want to make PostRank useless for the demographic that only has pageviews as a metric. And, as you’ve pointed out, that demographic includes both blogs that’re part of a community as well as completely obscure ones.
Yes, installing our widget will create a slight bias. It’s an extra metric to drive engagement analysis that not everyone chooses to have. However, again, most publishers can choose to install it if they want. In reality, though, pretty much any system will have demographics for whom their analysis is more favorable. We try to balance it as best we can, as does everyone else, but nothing’s perfect.
Ilya also let me know that we’ve recently updated and re-balanced our metrics and weighting, and views importance has actually been decreased. So while it’s not removing it as a metric entirely, it is a step in the direction you wanted. 🙂
Hi Melanie, well if it is PostRank and Ilya’s justification that it is there for those bloggers whose only demographic is pageviews as a metric then it isn’t an adequate reason. It is giving them a totally false sense of reader engagement and isn’t helping them with achieving their goal of building up their blog readership.
You have an amazing application that could be helping both bloggers and their readers. Yet because of the pageviews statistics it isn’t achieving its true potential.
As I said before (in my email) I like the concept of my readers having a quick mechanism that allows them to work out what of my posts are the “most popular”. But there is absolutely no way I will add your PostRank widget to my blogs when I know that the pageviews are flawing the metrics to the extent they are.
However PostRank remove this metric as part of their ranking system and I will proudly display your widgets on my blogs.
Now I know that making that statement you are probably thinking this is one blogger and it doesn’t really matter whether she does or doesn’t add it to her blog.
Normally I wouldn’t point this out, but sometimes it is important to make people aware since you probably aren’t. The Edublogger, is a blog that I author for Edublogs. Edublogs hosts over 300,000 blogs just on Edublogs plus lots of blogs on Edublogs Campus sites (we host a large number of Campus sites). This blog feed automatically feeds into every dashboard on our site. It has high readership, and site visits, by our users, educators and bloggers.
My role with Edublogs is to support educators in the use of technology especially with using blogs with their students. I strongly support and explain to our readers any technology that is beneficial for them or their students or helps them be better bloggers.
FYI – For those reading wondering what the pageviews are that we are discussing.
Pageviews are the metric that is displayed as an eye in your statistics on PostRank. They are calculated by the number of times a reader clicks on the title of a post in the PostRank widget.
I had never even heard of Post Rank before I read about it in this blog yesterday. Maybe I am simply not up to date with such technology. Are there other just as valid “popularity counters” out there? Has anybody got a list of them? Do they present the same sort of picture as at the top of this blog? If so, then they will tend to validate Larry’s list. If not, then it would suggest that it represents only a portion of the blogosphere.
My turn for a correction. I said “Pageviews or view stats are the metric that is displayed as an eye in your statistics on PostRank. They are calculated by the number of times a reader clicks on the title of a post in the PostRank widget” because based on my observations and how the widget appeared to be working that was my conclusion.
This is not correct.
The pageviews (represented by the eyes on a PostRank score of a post) are the number of times your post has been viewed. If you add a PostRank widget to your blog PostRank is able to monitor the number of views of your posts similar to how Google Analytics monitors visits to your posts. But if you don’t have a PostRank widget on your blog it can’t monitor this statistic.
Difference between Google Analytics and PostRank is Google Analytics is only ranking your posts on visits to your site whereas PostRank includes lots of statistics to determine how it ranks your post.
If you want to read more about the discussion of how PostRank is ranking posts and comparing blogs based on topics I suggests you check out the comments on my post here. Thanks to both Melanie and Jim for taking the time to explain so that we all get a better understanding of their metrics.
I also should add the following. Take for example my post on The Edublogger “Check Out This Two Page Blog Guide For Parents!”
This post has a current score of rank of 2.2 and most of the engagement making up that is 16 Delicious counts. But if you looks at Google Analytics the post has 726 pageviews for the month and ranks 7 in Analytics for the month.
Technically speaking, having the view stats in the ranking, is probably giving the post a more balanced metric.
So where am I at with adding PostRank widgets to my sidebar? Maybe — thinking about it.
If I can be given enough really good reasons of the benefits, of how PostRank widgets on student and class blogs will help them, then yes I will add them. Educators would love your thoughts on this — how might students be using this?
Followed a link back from our site and found your post– thanks for mentioning us!
Chaim Goodman-Strauss
mathfactor.uark.edu
Such a wonderful blog. Great job.