Further Thoughts On the Most Read Project

xerexes's picture

It's been two years since Comixpedia published an update to our Most Read project which tracked the audience shares of webcomics.

It was difficult to determine readership numbers then, it's just as difficult to conduct any kind of "Internet ratings" now. But it's an extremely useful process for Comixpedia as it helps to ensure that we are not overlooking significantly popular webcomics in our coverage (It is not a prerequisite that a webcomic be "popular" to merit coverage. The strength of readership of a particular webcomic, however, is a legitimate tool for deciding what we should write about). If you have suggestions for future efforts in this area feel free to post a comment here.

Comments

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
Ryan Q North's picture

Project Wonderful

One nice thing that Project Wonderful does is keep track of reader stats, so you can see and compare the hits across advertiser sites. Since a lot of our sites are comic sites right now, maybe this could be useful?
xerexes's picture

Yes

That is a very good idea.

____

Xaviar Xerexes

Mad, Bad and Dangerous to Gnaw.

Xaviar Xerexes Oh yeah... this place is called ComixTalk now.
joshl.'s picture

I don't know if this project

I don't know if this project is really a good idea, because being reminded of which webcomics are doing really well has a way of depressing everybody.
xerexes's picture

Hope Not

I would hope not. This isn't meant to say X comic is the best and all others are lesser. And it's pretty clear for example that Penny Arcade is (unless I find new sets of data that show otherwise) still the #1 webcomic on numbers. I suppose that might provoke bad feelings, but it's information that exists whether I write up the list or not.

If you're a creator and don't care about numbers I suppose it might be an article to avoid if for no other reason then you don't care about numbers. And if you're a creator and do care about numbers you probably already have a pretty good sense of where your comic is so I'm not sure this will tell you much. (Although I'm always interested in more data to correct flaws in the list)

And if you're not a creator then I'm not sure why you'd get emotionally involved unless you're an uber-fan of a specific comic.

Interestingly enough, Project Wonderful - for those webcomics using it - turns out to do exactly what I thought about writing up a script for in 2005. If more and more webcomics use Project Wonderful then there won't be much work to do in sorting out this kind of list b/c it'll all be in Ryan's database.

All kinds of media - blogs, video, etc on the internets are struggling with how to count what they do for advertisers. Thinking through that problem and exploring potential solutions is why I'm interested in this subject. It's an interesting topic for that reason alone.

____

Xaviar Xerexes

Mad, Bad and Dangerous to Gnaw.

Xaviar Xerexes Oh yeah... this place is called ComixTalk now.
timdemeter's picture

Man-Up.

joshl. wrote:
I don't know if this project is really a good idea, because being reminded of which webcomics are doing really well has a way of depressing everybody.

I'm thinking you're funning a little Josh, but for anyone who may take it for serious:

I know right where my comic sits in the grand scheme of webcomics, and am I perfectly happy with it? No. But ignoring that fact isn't going to help. Sure, some folk are in it just for grins, but some are not, and denying us data we can analyze and use to improve our mehtods of craft and outreach to spare some folks feelings is bad business and irresponsible for Comixpedia as a trade magazine. By your response I know you agree Xaviar, but I guess I'm more underscoring that this kinda idea tends to raise my ire. If we want webcomics to be treated like an industry, not comic book fantasy camp, then leave your sensitivities at the door, and get down to business becasue even great talent and good intentions will only take you so far.

Tim Demeter
does a buch of neato stuff.
GraphicSmash

Clickwheel
Reckless Life

Tim Demeter does a bunch of neato stuff. Clickwheel GraphicSmash Bustout Odds
EricMillikin's picture

Two thoughts

Two thoughts:

The common misconception that "traffic" or "hits" equals readership seems to still be floating around.

There are other options besides "treating webcomics like A) a fantasy camp or B) an industry." One of those options would be to treat it like art.

--------------------------

Fetus-X is the greatest comic in the world.

--------------------------

Fetus-X is the greatest comic in the world.

The William G's picture

I'm with the guy with the

I'm with the guy with the zombie arm on this.

If only because it'll eventually gets used in some dick-measuring contest anyway.


xerexes's picture

No Arguments With That

But it doesn't seem to me that all of this is in conflict with each other. To be fair about it most folks doing comics feel like they're making art, even those that also want to there to be commerce going on with that art.

But there's no reason for anyone not interested in commerce or numbers to be ashamed of making comics for comics (art's) sake. Most of the greatest artists in any medium you can think of didn't care about numbers.

But Comixpedia has a "trade paper" aspect to it and this is one of those topics that's interesting to me in ways that go beyond webcomics. If you are interested in commerce, in making some kind of business around making comics then unless you're going to go with the wealthy patron approach (grants would fall into this category too - I'd love to see more options for webcomics here - something like a Xeric grant for webcomics wouldn't be a bad start) the audience numbers matter. They matter whether you're going with a subscription, advertising or merchandise model.

And also selfishly for Comixpedia's coverage - a bunch of our coverage is what we (we being me and any other active editors and contributors - that changes over time) like but it seems fair and objective to me to check that with something like audience numbers. We may be missing something out there outside of personal taste that we should be writing about. I check that in other ways too by paying attention to awards like the WCCAs and other indicators from other groups.

____

Xaviar Xerexes

Mad, Bad and Dangerous to Gnaw.

Xaviar Xerexes Oh yeah... this place is called ComixTalk now.
EricMillikin's picture

Fair and objective

Xerexes writes: "it seems fair and objective to me to check that with something like audience numbers"


But you don't have audience numbers, right? So instead you have something neither fair nor objective -- a list of a small fraction of webcomics, limited by personal tastes, ranked by a series of systemic biases.


--------------------------

Fetus-X is the greatest comic in the world.
--------------------------

Fetus-X is the greatest comic in the world.

xerexes's picture

No

Eric

Maybe you're making a funny, but no you're not correct.

I have the same thing available to everyone else on the Internet (except for large corporations in industries covered by Nielsen Online and similar companies compiling reports which cost $$$. And which may or may not be anymore accurate depending on which analysts you want to believe). This consists of several options: data from a subset of Internet users collected by companies such as Alexa and Compete - this approach has flaws (which have been discussed on this site) but it is useful; data from publicly available counters such as Extreme, Hit Meter (and Project Wonderful), and data from private counters (log servers, stats programs). All of this is objective data. We can (and should) debate its validity and its flaws but that doesn't make it categorically "neither fair nor objective". (It's also not true that it's limited in any way by personal tastes or "systematic biases" as I've made (and keep working at) a tremendous effort to identify and incorporate webcomic titles that appear to have any significant readership at all. That has nothing to do with taste and is only limited by the limitations of publicly available tools to identify such webcomics.)

Measuring readership (or viewership or whatever the applicable term for any medium) is hard. Every other medium is having a debate right now on how to measure it (if they're not an established format/medium) or whether their methods for doing so are accurate (if they are established like TV, Radio, etc) so it's not surprising that any discussion surrounding webcomics is going to encounter a lot of questions and debate on the subject.

I understand some people for various reasons think considering the size of readerships to be (a) an affront to artistic purity and/or (b) an exercise that can never attain perfection and so is never worth the effort. I disagree on both counts and while I think I understand where those two points of view are coming from I just don't agree with them.

____

Xaviar Xerexes

Mad, Bad and Dangerous to Gnaw.

Xaviar Xerexes Oh yeah... this place is called ComixTalk now.
EricMillikin's picture

I mean numbers for the entire webcomics audience

I'm not saying you don't have some audience numbers, I'm saying you don't have enough audience numbers, and the numbers that you do have are crunched through sytems full of biases.

One of many examples would be the tendency to treat webcomics readeship as if it can be measured through the traffic on a site, when often the site features more than just comics, or the comics are featured on more than one site.

--------------------------

Fetus-X is the greatest comic in the world.

--------------------------

Fetus-X is the greatest comic in the world.

joshl.'s picture

I am, although I really do

I am, although I really do find such lists indicative of how little things tend to change.

The popular comics still always have the same qualities, with few exceptions.

"Consistency" - which means that the artists won't push their style or try to evolve, because the readers don't want them to.

"Innovation" - which means the writer has found various ways to make the comics while doing less work.

Lack of investment - which means it's easier to do well with a comic if the readers don't have to remember anything.

Catering to fetishes - Catering to various fetishes, such as anime.

All these things, and more, are still a lot more important than learning how to write and draw well.

Stop me if I'm derailing.

The William G's picture

The truthiness of this makes

The truthiness of this makes me weep.

timdemeter's picture

You're right Eric

You're right Eric, I left that out and I probably shouldn't have. I've always been a very commerical artist (you wouldn't believe the arguements I used to get into with some of my art profs in school) and I'll get wrapped up in that from time. I certainly don't want to come of as an us vs them kinda guy in terms of business vs. expression, that drives me crazy. Webcomics need both equally, and I do my best to support them both, even if I do tend to skew to the former.

Thanks for covering my ass there before someone really (rightly?) put me in my place, is what I'm saying.

Tim Demeter
does a buch of neato stuff.
GraphicSmash

Clickwheel
Reckless Life

Tim Demeter does a bunch of neato stuff. Clickwheel GraphicSmash Bustout Odds
scottstory's picture

Whether webcomics are

Whether webcomics are illustration or art is besides the point. A serious survey to discover what the webcomic demographic is is a good idea.

Do original hits equal popularity or readership? Clearly, that depends on the strip. Some are put out as a business proposition, monetized and marketed. Some are purely self-expression, and the point is not how many read it, but who gets it.

Like Tim, I fall into the commercial art camp, but that doesn't mean I think less of the fine art camp.

http://www.graphicsmash.com/comics/johnnysaturn.php

http://www.graphicsmash.com/comics/johnnysaturn.php
TCampbellWhoCan'tLogIn's picture

Josh's honesty is

Josh's honesty is refreshing, and it comes close to articulating what I feel is the core of resistance to information like this, and why so little is known despite the trackability of online work. As long as no one knows anybody else's numbers, it gives us a certain kind of equality that you'll never get in the numbers-obsessed motion picture business. That is nice and I will be sorry to lose it, but I would be more sorry to see a generation of artists continue to stumble around in the dark with no metrics to measure their audience against that of the field as a whole. Yes, if a reliable rating system were produced, it would be misused. It would also be used correctly, more often and by more influential people. Those of us who have serious ambitions for the long-term development of this form cannot do without this knowledge in the years to come. Xaviar, as it happens, I was just speaking with someone who could be of some help about this very topic. Please get in touch with me privately at your convenience. --TC
timdemeter's picture

Yeah, that!

TCampbellWhoCan'tLogIn wrote:
Yes, if a reliable rating system were produced, it would be misused. It would also be used correctly, more often and by more influential people. Those of us who have serious ambitions for the long-term development of this form cannot do without this knowledge in the years to come.

This is totally what I meant in much better words. Much to learn, still have I.

Tim Demeter
does a buch of neato stuff.
GraphicSmash

Clickwheel
Reckless Life

Tim Demeter does a bunch of neato stuff. Clickwheel GraphicSmash Bustout Odds
EricMillikin's picture

What?

Campbell writes: "Those of us who have serious ambitions for the long-term development of this form cannot do without this knowledge in the years to come."


Did you just divide this discussion into "Those who have serious ambitions for the long-term development of this art form" and those who don't? That's pretty much a dick move.


And I think you're also going to have to make a better case for the idea that, after several millenia of existence on this planet, the human race has somehow not yet begun to get serious about the arts because we don't have a good system of hit tracking yet.



--------------------------

Fetus-X is the greatest comic in the world.
--------------------------

Fetus-X is the greatest comic in the world.

TCampbellWhoCan'tLogIn's picture

Eric, I tend to assume that

Eric, I tend to assume that everyone in this discussion does care about webcomics' long-term future. Above, you say: "But you don't have audience numbers, right? So instead you have something... limited by personal tastes, ranked by a series of systemic biases." I agree with this completely, and I think Xaviar does too, which is why he's reopening the idea of a traffic ranking system. (Knowing Xaviar as I do, I think he's making a huge sacrifice on the field's behalf.) Such a system, at the very least, introduces some biases we don't already hold. If a study reveals that poker comics are starting to approach gamer comics in audience size, then that should probably be reflected in coverage. And cartoonists considering the next move in their career might be interested in those numbers, if they have a genuine interest in poker. But what about those Iconclasts who could care less about such big numbers? Aren't they part of "people who care about...?" Well, take Harvey Pekar. Even after the movie, I don't think too many people look at Harvey and think "sell-out." His direct market numbers will never approach Superman's and his bookstore numbers will never approach Naruto's. But Harvey most definitely had to look at his place in the marketplace when deciding how many issues to put out next year, how much he could afford to pay artists and what his long-term plans should be, plans which ultimately led to the production of more Pekar works than we'd otherwise get. We have "hit tracking" for comic books and bookstore sales and any art form that has enough professional practicioners to be called an "industry." I understand you when you say that you would prefer to regard webcomics as art, not as an industry. But we live in a world where everything is connected to everything else, and artists need to understand how many people their art is reaching so they can get on with the process of feeding themselves and producing more art, as effectively as they can. HOW aware they should be of this varies with the artist, and I think we'd agree that if they're thinking of their numbers as they produce the art itself, that's a problem. But the information should be there to help them as they make life decisions. I tend to believe that in the crushing majority of cases, and certainly in this one, information is preferable to ignorance. By saying this, I'm not trying to paint you or Josh as someone who prefers ignorance to information. I am trying to put forth my thoughts in such a manner that might lead you and others to agree with them. We'll see how I do.
EricMillikin's picture

More on "hit tracking"

Campbell writes: We have "hit tracking" for comic books and bookstore sales and any art form that has enough professional practicioners to be called an "industry."

Sure, and those tracking methods are generally very clear and up front about their limitations. We're very careful to refer to sales as sales, readership as readership, circulation as circulation, etc. Bookstores don't take "traffic" and equate it with "readership," for example.

I think there's just too many comics off our radar, too little data on those that are on our radar, and what data there is ends up getting misinterpretted and misused.

Oh, and sorry if I misinterpretted your "Those of us that are serious about the future" or whatever as more divisive than you intended.

--------------------------

Fetus-X is the greatest comic in the world.

--------------------------

Fetus-X is the greatest comic in the world.

timdemeter's picture

We're only talking data here.

Not to put words in T's typing fingers but I think he and I are kinda trying to make the same point here, (imagine that!) and it's not so much about the artistic ambitions of the medium or creating some kind of divide between art and business, we're simply saying that from the more business focused point of view we're taking on this issue, numbers can be an important thing to distect. It is equally important that boundries be pressed both on the creative and financial frontiers, but not everyone is best suited to push both. Personally, I know I'm better suited to the former, and leave the latter to those best suited in that arena.

In short, you're right this data should neither refute nor validate the artistic credibility of any given piece in any way, and I don't believe either of us are saying that, but it can help define trends that may be valuable when looking at things from the other direction and I see no harm it may casue to any sort artistis pursuits. Sure, it may raise that popularity contest thing that some may find distasteful, but don't we spending our time looking at webcomics as an industry have as much right to pursue our ends as those looking to further it artisically?

Not to sound all Johnny Peacemaker, but I think we're all just getting lost in well-meant semantics here.

Tim Demeter
does a buch of neato stuff.
GraphicSmash

Clickwheel
Reckless Life

Tim Demeter does a bunch of neato stuff. Clickwheel GraphicSmash Bustout Odds
timdemeter's picture

Okay we posted that at, like, the exact same time.

So I'll add:

What T said.

Tim Demeter
does a buch of neato stuff.
GraphicSmash

Clickwheel
Reckless Life

Tim Demeter does a bunch of neato stuff. Clickwheel GraphicSmash Bustout Odds
EricMillikin's picture

If you want to talk business, talk business

I don't quite get that either, Tim -- if you want to talk business, then why not talk about business? You know, dollars and cents rather than hits and page views.

To me talk about traffic just gets into weird numbers games, where the page views of sites with thousands of pages get compared to sites with few pages, or the daily unique visitors of a daily comic get compared to the daily unique vistors of a monthly comic, or whatever.

--------------------------

Fetus-X is the greatest comic in the world.

--------------------------

Fetus-X is the greatest comic in the world.

timdemeter's picture

Crap, now I have to actually think.

EricMillikin wrote:

I don't quite get that either, Tim -- if you want to talk business, then why not talk about business? You know, dollars and cents rather than hits and page views.

In a lot of ways, especially with ad sales the seperation gets murky, was my thinking.

EricMillikin wrote:

To me talk about traffic just gets into weird numbers games, where the page views of sites with thousands of pages get compared to sites with few pages, or the daily unique visitors of a daily comic get compared to the daily unique vistors of a monthly comic, or whatever.

That's a good point, I don't have a good answer for. I guess in using any numerical data to it's fullest the context does not to be taken into account as well. If there's a mechincal way to measure that it's well beyond me, and we'll just have to take any hard numbers with a grain of salt and have to put some elbow grease to making the numbers make as much sense as we can within those contexts.

OR: Ryan invent something to solve this, pls/thx.

As much as I'd love to debate this all evening it is the party night amongst party nights, and I'm going to go search for the answer to this at the bottom of a beer bottle. (If I don't find it in the first one, don't worry, I'll try again.)

Happy Turkey day everybody.

Tim Demeter
does a buch of neato stuff.
GraphicSmash

Clickwheel
Reckless Life

Tim Demeter does a bunch of neato stuff. Clickwheel GraphicSmash Bustout Odds
EricMillikin's picture

Yeah, Happy Turkey Day everyone!

Yeah, Happy Turkey Day everyone!

See you at the country club!

--------------------------

Fetus-X is the greatest comic in the world.

--------------------------

Fetus-X is the greatest comic in the world.

xerexes's picture

Maybe...

Eric,

I replied to your other comment above. I'll stay out of your discussion with T though.

As for myself let me reply only to your last sentence there. Taking art seriously is an apple to a measuring the readership orange. There have absolutely nothing to do with each other in the grand scheme of history or human experience, other then the coincidence that sometimes great art does attract a mass audience. Sometimes it doesn't. But in terms of coverage today - a mass audience for something is in and of itself can be newsworthy. And again, just b/c something doesn't have a mass audience doesn't mean it's disqualified from coverage.

And really I have no idea how measuring the audience for webcomics in anyway is a comment on the nobility, seriousness and universality of the human experience of art. How do we get from looking at data on unique visitors, page views and other server data to debating whether or not the human race is serious about art?

To put my point in immediate perspective I think Fetus-X is a great, boundary-pushing comic for which I'm not going to change my own opinion about if I found out it was rivaling Penny Arcade in readership numbers or at the other end was only read by a small number of people. And I hope that we'll catch and write about the next Fetus-X out there, but we're going to do that by the hard work of finding it ourselves - not b/c of its readership numbers (in fact hopefully when we find it we can send new readers to check it out). There's no reason in my mind, however why that subjective effort to cover quality work has to conflict with this more data-driven project.

____

Xaviar Xerexes

Mad, Bad and Dangerous to Gnaw.

Xaviar Xerexes Oh yeah... this place is called ComixTalk now.
EricMillikin's picture

Serious about the form

Xerexes writes: "I have no idea how measuring the audience for webcomics in anyway is a comment on the nobility, seriousness and universality of the human experience of art"

Yeah, me either, that why it seemed really weird to me that Campbell wrote "Those of us who have serious ambitions for the long-term development of this form cannot do without this knowledge in the years to come."

I think maybe he meant something else more specific maybe.

--------------------------

Fetus-X is the greatest comic in the world.

--------------------------

Fetus-X is the greatest comic in the world.

Zach's picture

I'M WITH ERIC!

The knowledge of how many people are reading our comics will ruin everything! Look how the popularity ISBNs on books has ruined writing! And I'm completely against having trackers. They're utterly biased. I mean, Ryan's tracker is several percentage points different from my two other trackers occasionally. And they don't track real data anyway - just IP addresses. What about people thinking about my site, animals looking at it, or people whose alphabet soup occasionally spells out "smbc?" Or what about people who appreciate my work on a more sublime, abstract, artistic level? They should count at LEAST three times. And cross-referencing sites against each other would be even worse! How can I still be an artist after checking (nay, being FORCED to check) other people's stats? I might find out they're getting more readers and assume they're better then me. Then, compelled by these unreliable numbers, I would cease making single panel romance comics (art), and start doing a strip with broad appeal (FILTH). Of course, the true terror her is not in the technology itself, but if it gets loose! What if a new stat checker is created, but then comes alive, leaps off the table, and begins ravaging the countryside? Tonight, I'm gathering the townsfolk so we can storm Project Wonderful! Everybody Wins, you say? Tonight, everybody BURNS.
joeymanley's picture

ISBN's don't have anything

ISBN's don't have anything to do with popularity.

Joey
www.webcomicsnation.com

ubersoft's picture

This isn't *conceptually* a bad idea, but...

...as many have said, there is no reliable way of doing it without getting ahold of every webcomic's server stats.

Alexa is a *deeply flawed* method for measuring site traffic, on MANY levels. And using a crappy methodology just because you can't get access to a better methodology doesn't make the ratings any more accurate -- it's still exactly the same level of crappy.

xerexes's picture

Imperfect Yes but...

I'll grant that there are issues with Alexa's approach, but "deeply flawed" on "MANY levels"?

It would be helpful to spell out for us what pushes it beyond imperfect to "deeply flawed" and identify all of those levels you're talking about. Write it here, or point me to a URL of something that explains your conclusion - I'd like to hear it.

 

____

Xaviar Xerexes

Mad, Bad and Dangerous to Gnaw.

Xaviar Xerexes Oh yeah... this place is called ComixTalk now.
The William G's picture

It's because toolbars are

It's because toolbars are the work of Satan. I've been saying this for years now.

ubersoft's picture

OK, here is why it is deeply flawed:

1. Alexa must be downloaded and installed on a browser in order for it to "count." This automatically skews the sampling of the web-using population, because it only records the web browsing movements of people who are willing to install spyware on their browser.

2. Alexa is not available for any browser other than Microsoft Internet Explorer. Taken as a whole, when you look at the entire population of the web, this doesn't look like a significant statistical outlier, because IE comprises 75-80% of the entire browser market, but in truth it is a VERY significant outlier on some sites, because general use trends do not correlate to the usage trends of specific sites. For example, according to my server logs only 25% of my audience uses Internet Explorer. This means that even if all other things were equal, only 25% of my audience is being counted by Alexa. The first item is debatable, I suppose -- not everyone considers these things "spyware" to the degree that I do. The second, however, is what makes Alexa so flawed. In order for Alexa to work it has to assume that every website it counts reflects the usage trends of the web as a whole -- in other words, that 75% of visitors to every web site will be using Alexa. That is simply not true.
joeymanley's picture

Add Alexa's flaws to the

Add Alexa's flaws to the methodological problem on Comixpedia's end that any list they come up with can't possibly include all webcomics -- or even all popular webcomics (to this day, I'm still finding comics in weird little niches that apparently have decent to large followings, and are well-known in certain circles, but are never talked about in webcomics circles).

So: incomplete and flawed data on an incomplete and flawed subset of the phenomenon being researched.

That's pretty much the definition of GIGO.

I'm not meaning to trash-talk Xerexes or the project in question here. Just pointing out that there are all kinds of reasons to be cynical about this, not just ideological ones.

Joey
www.webcomicsnation.com

EricMillikin's picture

Alexa lists 6 known biases

The issues with Alexa traffic rankings came up as a tangent on a convention panel with Rob Balder, Howard Taylor, The Ferret and I last spring. Alexa lists 6 known biases (and says their biases "are likely not limited" to those six) and also points out that their relatively small sample size makes it difficult to distinguish between the traffic rank for sites the further you get from the highest trafffic sites.


Those six biases are listed at the bottom of http://www.alexa.com/site/help/traffic_learn_more


They include some of the things ubersoft mentions: Biases toward certain browsers, certain operating systems, certain countries, certain sites that may share content, certain sites that Alexa users are more likely to visit, and against sites with secure pages.


--------------------------

Fetus-X is the greatest comic in the world.
--------------------------

Fetus-X is the greatest comic in the world.

joeymanley's picture

Comscore.com is generally

Comscore.com is generally considered as the best traffic measurement system -- at least, according to three different sets of business people I've had meetings with lately. I don't know why those people thought this about Comscore, though. Maybe they know something about their methodology I don't. Or maybe it's just because Comscore costs a lot of money, and these guys are paying for it, so they need to believe. Dunno.

I do know that WCN vs. Comic Genesis looks very different in Comscore than in Alexa. "Very different" as in "hard to believe these two systems are supposedly looking at the same two websites' traffic."

For my own ego purposes, I prefer the Alexa data in that particular case, by the way. Ha!

Penny-Arcade rules all our asses on Comscore just as handily as they do on Alexa, by the way.

Joey
www.webcomicsnation.com

pclips's picture

They're not

Alexa fucks up Comic Genesis data bigtime, Joey. This is due to a bias they do NOT list.

It gives separate ranks to separate servers. So what you see for the Alexa rank of a given Comic Genesis comic is almost a random number (way too high for the comic's traffic, but way too low for all of Comic Genesis) based on what other comics that one happens to be grouped with on the server.

Because of this problem, certain webcomics get completely rooked in Alexa rank because of how they handle their servers. For example, Sluggy.com (the comic's server, rank 13,049) does not get the traffic counted from Sluggy.net (the forums server, traffic rank 73,687). There is no easy way to combine those rank numbers to get an accurate feel for Sluggy Freelance's traffic relative to a comic which hosts its archives and forums on the same server.

C+A+D at http://www.ctrlaltdel-online.com/ has a rank of 3,616, which does not count traffic from http://www.cad-comic.com/ which is still a whopping 22,381.

This is why a comparison between any two comics may be flawed. You have to dig and investigate to make sure Alexa is comparing apples to apples.

It is better to use Alexa data to try and learn something about patterns of growth among properties and ask broader questions.

pclips's picture

Comscore

Wow, at first blush, it looks a lot like Comscore has got a real scam going. Their claims are outlandish and unsupported. I'll have to do some more research. Maybe I can find a good writeup on them. But my instincts scream "this is bullshit" right now.
joeymanley's picture

Well, it appears that

Well, it appears that Comscore owns/is the latest incarnation of Media Metrix, a brand I am very familiar with from my days at StreamingMedia.com and Streaming Media Magazine. Advertising industry executives take Media Metrix as seriously as they do the Nielsen Ratings. Which is to say, they know the numbers are not 100% accurate, but they seem to believe it's the best they've got. Why they seem to believe that, I don't know. They probably know more than any of us here on this message board. But it's always possible that they don't. And it's also always possible that my impression was incorrect.

The people who are pushing Comscore on me are well-known, highly respected, and very successful business people in the field of venture funding. But, again, that doesn't necessarily mean anything, either.

If you find anything on them, let us know.

Here's their Wikipedia page. It has links to stories about them on CNET, MSNBC.com, and so on, some laudatory, some not.

Joey
www.webcomicsnation.com

pclips's picture

Good point, Eric.

However, there is useful data which can be gleaned even from a flawed system of measure. For example, Alexa indicating that schlockmercenary.com has higher traffic than partiallyclips.com is a datum you can have near 100 percent confidence in.

If you use it primarily as a relative measure of traffic, and consider groups of comics within certain bands and ranges, you can learn a lot.

Alexa's biases are frequently misunderstood and overstated, and tend to apply mainly to sites with traffic above 100,000 rank, where Alexa does not even pretend to accuracy. The relative rank of 2 sites with a greater than 10 percent rank difference is strongly correlated to other traffic measures, at least within the top, say, 25,000 sites.

The concept of imperfect information being better than no information might be counterintuitive. People love to discredit sources by catching them in some inconsistencies. But the difference between Alexa data and reality is not a whole lot greater than the difference between poll data and election results. Polls are imperfect, often misleading, but generally useful tools.

I class Alexa the same way. You have to study it and understand what you are looking at, account for biases within reasonable ranges, and know the degree of uncertainty in the data. You can gauge its accuracy by trying to predict future trends and tracking the results. In the end, it is a useful tool...just a tricky one.

joeymanley's picture

Agreed. I've been putting

Suggestion: maybe another way to get at popularity is to do it the way Technorati does it, at a remove: rather than rank blogs according to an attempt to discover actual readers (which is difficult to pin down in an RSS-heavy environment like blogs), they rank blogs according to how many links there are in other blogs -- which indirectly measures the size (because the people reading are the ones making the links) and enthusiasm of the audience.

Many webcomics -- the ones that offer RSS feeds, anyway -- show up in Technorati's ranking as if they were blogs. Many others do not. For example, WCN looks like a vast blog to Technorati (it's ranked 2,438 at the moment, with 3,376 links from 737 blogs), but Drunkduck and Comic Genesis do not.

[oops -- accidentally overwrote my own earlier post in an attempt to reply to myself -- damn! It was a long-winded Manley(tm) post, too. Crappity crap crap crap. Will try to recreate it later).]

Joey
www.webcomicsnation.com

joeymanley's picture

Here's my attempt at

Here's my attempt at recreation.

On the subject of Alexa's usefulness: I agree with pclips that it has many powerful uses, one of which is to get a broad sense of how traffic moves across the, um, webcomicosphere (a word which I hate, even though I just, as far as I know, coined it). I've been doing a lot of data mining, as has Rob, I think, in precisely this way. But when it comes to ranking individual comics against one another in a chart, the problems seem more difficult to justify, to me, than they do to Rob. Here are some more (and I'm focusing on WCN comics not out of a sense of entitlement or whatever, but just because I'm most familiar with the issues faced by WCN comics vis a vis Alexa -- I'm sure other sites have other issues, each probably unique to its own circumstances):

WCN comics all look like WCN to Alexa -- they're not split out separately. I have the data, but am not sure if I'm allowed to share it (it's something the artists would have to give me permission to do).

Like Sluggy, all WCN comics have their forums housed elsewhere, some at TAC, some at even smaller sites. This puts WCN comics at a disadvantage compared to other sites where the forums are at the same URL. Likewise, the creator's blogs, when they exist, are housed elsewhere. Again, a site with an active blog plus a comic plus a forum will have an advantage in Alexa, which doesn't necessarily reflect the true size of the audience for the comic itself.

Many of our comics are actually more popular in the form of syndicated feeds (embed-ready tooncasts or pure RSS) than they are on WCN itself. For example, American Elf is syndicated on the comics page of a Malaysian newspaper with millions of readers (or, at least, with an Alexa ranking in the three-digit range) -- how many of those readers click through to the comics page and then to American Elf is an open question, only known to the webmaster of that Malaysian website. Now factor in the dozens of other places American Elf is syndicated. Now factor in all the comics on WCN with syndication feeds. RSS itself can be consumed outside of the context of the web -- for example, some number of people suck down the various RSS feeds on WCN for consumption within desktop applications like Outlook or Sharpreader, or on handheld devices like the PSP (where they don't even have to be connected to the Internet to read the comic, in cases where the comic includes an enclosure). I have data for the RSS feed itself, but there's currently no way for me to know how many individual users are reflected in a request for the RSS feed implies (bloglines pings the feed once every thirty minutes, but then mirrors the contents of the feed to all its subscribers based on what it picked up in that one ping, while desktop aggregators might ping fifty times before the individual looks at the feed even once -- and so on, and so on). RSS and tooncasting aren't extremely widespread in webcomics right now, but they will be more and more as time goes on -- and our RSS audience seems to be exceptionally loyal and enthusiastic, based on anecdotal evidence -- essentially, that's where we're building our new generation of hardcore fans, all of whom are hidden to Alexa. I personally don't show up in the server logs of Name Removed -- a great not-very-popular webcomic that surely deserves my "Alexa vote" -- because I read the official full-content feed in SharpReader.

That's not to even bring up the issue of unauthorized feeds for popular comics out there.

And there was more, but I've forgotten most of what I posted earlier. Sorry.

Joey
www.webcomicsnation.com

pclips's picture

I agree that Alexa screws

I agree that Alexa screws all sites which host multiple comics on the same server, or split related traffic between servers. I'm sure that causes you no end of annoyance, Joey.

Those comics need to use some public tracking tool like Extreme Tracking or something, to get data which compares to standalone comics. I also agree that it does not count the fast-growing segment of RSS feeds, Clickwheel, and other non-web methods of electronic syndication.

There are many, many things not measurable by Alexa to any degree of accuracy. But many, many other things can be. And it bugs me when people dismiss the whole tool as useless when it is useless mainly for the things they want to measure.

If this were my project, I would do a literal ranking of comics against one another for only the top ten by traffic. I would then list a group of about the next twenty, the twenty after that, the fifty after that, and "candidates" for the 200 after that. Accuracy for inclusion by group is going to slide as you go down the ranks list, but reliable figures should be obtainable for categories that broad, by correlating multiple sources.

joeymanley's picture

Correlating Multiple Sources

Correlating multiple sources, eh? Okay, let's try it, using the most popular comic known to us (given that we all agree that the more popular a website is, the more accurate the data is likely to be -- and the more likely it is, presumably, that most sources will be in sync).

Right?

Here's Penny-Arcade.com on Alexa for the last year:

Here's Penny-Arcade.com on Compete for the last year:

Compete is another tool-bar based service, mentioned first to me by one of my new friends in the venture funding world, and apparently more trusted in that world -- or at least the part of that world where I've been hanging out -- than Alexa, for whatever reason.

So, yeah. P-A is either maintaining and building (Alexa) or falling like a rock (Compete). Care to correlate? I feel like I know less now, after looking at two different sources, rather than more.

Personally, I don't believe either set of numbers. All I know is that P-A is astoundingly popular, and the only reason I really know that is, well, they've linked to one of my websites before, and I've seen with my own eyes what happens. Beyond that, I have no real data.

Um, but, you know, for ego reasons, I prefer Alexa's numbers over Compete's for WCN -- but ego reasons are not really valid reasons for any purposes except for, well, ego purposes. Ha!

Alexa's good for watching general trends (and I mean VERY general trends -- like the way that comicbookresources.com, newsarama.com, and dccomics.com all have exactly the same-shaped curves over the past four years -- with several weirdly mirrored peaks and troughs all at the same times -- implying, to me, anyway, that their respective readerships have almost 100% overlap, and maybe even represent All of Print Comics Fans Online). For comparing individual website to individual website in an attempt to see who happens to be how much more popular than who else, in any fine-grained way (monthly, for example, being too fine-grained, to my tastes, for Alexa data) -- no. Not for me. But that's just me. I totally don't blame anybody for wanting to use it, or believe it.

[EDITED TO CORRECT GRAPHS -- I was originally showing P-A on Alexa for one month, compared to P-A on compete for one year -- corrected version shows apples-to-apples (one year to one year) comparison instead.]

pclips's picture

re: Correlating Multiple Sources

It's a good start. The more data points, the better, though. I think to tackle the question, you'd want a broad swath of data. Here's my ideal case:

Take three comics which have resided in the top 50 for traffic for at least the last year. Pick three which use Extreme Tracker as well. Get a three month snapshot of their Alexa rank, Alexa reach, Extreme Tracker monthly uniques, Complete tracker stats (whatever that's showing--I'll have to look at Complete more closely), and (this is the tough part) whatever real server stats you can get from the creators themselves (preferably two distinct metrics such as pageviews and uniques for each of the last 3 months).

The point is to evaluate the tools here, so you'd want three comics with as close to the same hosting setup as possible. I think I can arrange an analysis like this but I'm pressed for time, since I am launching a new comic at the end of the week. :)

joeymanley's picture

My research is for a private

My research is for a private business document -- a formal business plan to show to potential investors -- so I doubt I could get the level of cooperation from cartoonists outside of my business that would cause them to give me their private server logs. And rightly so. I wouldn't share my private business data with somebody just looking to write a business plan, either, especially somebody who might be seen as my "competitor" (though that word doesn't really mean squat in webcomics, where all of us are working together, in a way, to build larger audiences and bigger opportunities).

That said, the info you've provided here has been helpful, and I hope that anybody else who is interested in this kind of stuff has found our public discussion helpful, too. And I wish you luck with your own research project.

Thanks.

Joey
www.webcomicsnation.com

xerexes's picture

It really has been helpful

This is exactly the kind of discussion I was interested in learning from and thanks to all of the readers who contributed to this thread. I still have some hope to craft something interesting and useful although it is going to take some coding to pull off I suppose.

Primarily I'm going to need to write some kind of bot to pull what is publicly available (public stats, Project Wonderful) and what others might provide to me from their server stats (of which I think I could agree to keep actual numbers confidential - using the data only to rank the site). It seems to me that if I could pull such data into a db I'd have a healthy data set based on page views. We know that's not going to be represent all of the varying ways comics get through the "internets" to people - there are other protocols, etc. but it'll be what it is. (I've always tried to list all of the exceptions known about the lists when I write them up and this would be no different.)

We could then use the relationship of the sites in this database to make some educated guess about where sites not in the db fall. At present I suppose that's going to rely on sources like Alexa, Compete, etc. But the more sites I could get into this proposed DB itself the better the list could be I think.

Thoughts? Volunteer coders? :)

____

Xaviar Xerexes

Mad, Bad and Dangerous to Gnaw.

Xaviar Xerexes Oh yeah... this place is called ComixTalk now.
pclips's picture

Competitors

Yes, business plans require you to name and research your competitors. Just the act of doing that has subtly changed my outlook on the webcomics community a bit. It's a nasty little virus, that old-school approach to IP.

But certainly, if we begin to think of one another as competitors, that is what we will become.

joeymanley's picture

Alexa Can be Injurious to Your Wealth

Today on GigaOm, an Internet industry blog, there was a post that made me think of this conversation. An excerpt:
The point is not to pick on Alexa, but to bring into focus the biggest shortcoming in the post 1990s-web: lack of a good dependable yardstick for ranking websites and web services. As more and more web properties come into existence, it is time for the industry to develop a more dependable, and open source tool to track general traffic trends, and web site rankings.

[...]

A non-partisan group could collect the data; much in the way open source projects keeps track of their code. There will be privacy and other related issues, but then these companies are chockfull of smart guys with all the answers.

The full post is here