This site now averages enough monthly traffic that it was eligible to become part of a CPM banner ad network. As you can see, I’ve now rolled out some ads.
The nice thing about a CPM banner networks is that you get paid per page impression (CPM) rather than per click. Since I’m getting paid per page view, I can actually estimate what I’m likely to make. That single banner will likely yield more than enough money to cover my server costs (and the server costs for youngphilosophers.org and sympoze.com). If high paying campaigns end up running on the site, then I might actually have some leftover to fill in 25% travel budget cut that was leveled on SUNY-Fredonia faculty (that’s a story for another time).
I don’t know of many philosophers that have tried to monetize their sites in this way. I’m not sure what the reason is, but I suspect that it’s one of the following two. They either think they don’t have the traffic to make it worth trying to generate ad revenue, or they think ads are annoying and don’t want them on their site.
With respect to the first claims, it may well be that a lot of philosophers are wrong about this. I didn’t think my site would be able to generate enough revenue to make it worth my time, but now I’m starting to think that I was wrong about this. We’ll have to wait and see if my estimates pan out. I’ll post a little review about all of the things I’ve experimented with in a couple of months, and you’ll be able to make projections for your own site.
With respect to the ads-are-annoying reason. I agree they can be annoying, but I think that annoyance might be worth the cost. In the remainder of this post, I’ll explain why.
Why Philosophers Should Consider Monetizing Their Sites
- It’s money
That’s the obvious reason. Even if you don’t think you could make much. I suspect a blog with a decent amount of traffic could generate enough revenue to give yourself some money for travel/research. If you don’t get a sweet travel/research budget, that’s nothing to balk at. And if you generate Brian Leiter traffic, you could pull in an extra $10,000 a year (I can walk you through where I got that estimate some other time). - The freebie sites are making money from your work by advertising anyway.
So you’ve found a free site that let’s you have a webpage or blog. Great. But, they’re going to make money off of you by advertising. So if you had some, general, aversion to advertisements. It’s worth noting that most of the free services you use generate money with ad revenue, and even if they don’t put ads on your site, they reserve the right too. (check the wordpress.com user agreeement). So you contribute to a company that produces annoying ads and serves them to your readers. Why not do it yourself. You have more control over the ads, and the money will be going to you which you can put to better use than the companies that are getting fat off your labor. For example… - Funding for other projects to benefit the profession
I currently maintain Sympoze and Young Philosophers because I think they are both great things for the profession to have. Right now, I pay for the services out of my pocket. Something that I hope will change soon. But in addition to covering server costs, I’d like to pay developers to add some features to sympoze. I’m also worried that the funding won’t always be there for young philosophers (In fact, I’m worried that it will run out next year). Monetizing these sites (in addition to this one) could handle some of these worries. - Our Profession Already Embraces Advertising.
The journals you publish in, and our main professional organizations all generate ad revenue. So again, you contribute time and effort to organizations that make money off of your labor by displaying those annoying ads.
Bottomline: The philosophy blogosphere generates a lot of traffic. There is revenue to be gleaned from it that can be put to good use for the profession. More philosophers should look into it.
Another Way the Profession Can Benefit from Ad Revenue: Open Access
As some of you know, there was a period of time where I ranted about open access philosophy…alot. You can browse through all of those rants here. In this post, from that series of rants, I outlined a way for fixing the peer-review issues everyone’s been experiencing with an open-access journal that generated ad-revenue. In this post, from that series of rants, the costs of running a journal came up as an obstacle to an open-access journal. Again, I suggested that a high quality open-access journal could probably off-set the costs by generating ad revenue.
So, part of my reason for experimenting with advertisements is to further test the feasibility of generating ad revenue for an open-access journal. Yeah, I’ve been thinking about this since last year. Here’s the procedure for figuring this out.
- Figure out how much a site like mine can generate (which I suspect won’t be terribly much, but also nothing to balk at).
- Compare my traffic to the traffic of one of the current high-quality open access philosophy journals (e.g. Philosophers’ Imprint).
- Estimate how much that journal could generate with a CPM network.
- Compare that number to the numbers thrown out here. This would give us some idea as to how well an open-access journal could cover operating expenses.
Incidentally, I contacted David Velleman back when I first started thinking about this to see if he had traffice stats for Philosophers’ Imprint. He said they were having some issues with their tracking, and that I should check back in a few months. I’ll be doing that soon. I imagine it’s WAY higher than this site. So, I’m optomistic that my idea above is feasible. We’ll have to wait and see.
I’d LOVE, LOVE, LOVE to see an open-access (high quality) competitor to Analysis, and I really think advertising revenue is going to be the way to go. Give it a couple of months, and I’ll have a better sense as to how feasible this is.
Conclusion
I know advertising on a site can be annoying. There’s always something a bit refreshing about philosophy blogs with nice, clean templates that are ad free. But if philosophers can be less dependent on their adminstrators for travel funds and other projects to benefit the profession, and if the profession as a whole can avoid being held hostage to big publishing companies that charge way too much for a business model that’s clearly outdated, then I think it’s worth that minor annoyance.
Hey, Andrew —
I’ll check with the folks at Michigan’s Scholarly Publishing Office (SPO), who publish the Imprint. I don’t know whether they’ve ironed out the kinks in their statistics software.
I am quite sure that the reason why open access hasn’t spread as widely as we had hoped is the lack of a “business model”. SPO invests a fair amount in the production of Philosophers’ Imprint; Steve Darwall, Victor Caston, and I do a fair amount that would be done at other journals by paid staff; and this year we have started investing some of our personal research funds in the operation, as the volume of submissions has increased. I don’t think that we would want to take advertising — not yet, at least. I’d better not make predictions about what we’ll do if submissions continue to increase at the current rate.
One caveat: I believe that a CPM network will place third-party cookies on a user’s’ machine and then track which other ads, at which other sites, the user views. The network thereby compiles a profile of the user’s interests, which is used in auctioning advertising space. (Auctions are held in real time whenever a user accesses a page. Advertisers “bid” for the advertising space on the basis of the user’s likelihood of being interested in their ads.) Maybe there are networks that don’t do this. You probably know much more than I about it. In any case, this practice raises very serious issues about online privacy and anonymity. So we would want to be careful about which networks we got tangled up with.
Thanks David,
That’s a very good point about privacy. The ad network has an option called “Participate in Behavioral Targeting” that publishers can opt out of. I opted out of that, so anything that shows up on my site is not supposed to be part of their Behavioral Targeting campaigning. But now that you mention it, I’m going to comb through their TOS more carefully double check that this opt out eliminates all cookies.
It would be nice to know which networks are more conscientious about this. This will be a good place to start. I’ll post more later.
Hello again, Andrew —
I have now tried to decode our web statistics. Keeping stats on a journal is not as simple as keeping them on a single web page, but here goes.
In the first half of 2009 (January through June), there were almost exactly 10,000 visitors who browsed the journal from its homepage, by clicking “Browse” or “Contents”. There were also 2800 visits to the “list of all pages” for individual articles (see below for what that means). And there were 200 searches.
Now here is the complicated part. If you called up an article via its own URL (say, http://www.philosophersimprint.org/009007/ for Stephen Finlay’s article on internal reasons), then you bypassed the journal’s homepage and weren’t counted among the 10,000; instead, you called up the article’s “list of all pages” and were counted among the 2800. But if you accessed the article via the journal’s “Browse” or “Contents” views, then you almost certainly bypassed its “list of all pages”, which isn’t directly accessible from those views. You almost certainly bypassed it, but not quite certainly. For with a couple of clicks, you can always get to the “list of all pages” for an article. So there may be some overlap between the 10,000 and the 2800 — people who browsed for an article and then navigated through to its “list of all pages”. And of course the 10,000 represents browsers of the journal, some of whom may not have accessed any articles, others of whom may have accessed several.
I guess you can think of it like this. If PI were a print journal, then in the past six months:
— 10,000 would have picked it up and scanned its table of contents, with some but not all going on to read one or more articles.
— Somewhere around 2500 would have gone to the journal in order to get a specific article.
— Another 200 would have looked in the journal’s Index for authors or topics.
Alles klar?
Oh — they’re not necessarily unique visitors. So for “people” read “person-stages”.
David,
Thanks for looking into this. Some thoughts…
First, those numbers seemed low to me. I would have predicted higher numbers. Then I realized that they do not seem to be track records for the whole site. For example, the number doesn’t seem to track
a. visits to the main page – http://www.philosophersimprint.org
b. visits to the About page (where philosophers would go to find out how to submit)
c. visits to the submission page (which is located off-site)
I bet there are people who visit those pages that are not counted in your 10,000. It’d be interesting to know what the number is if you include the main page, the about page, and the submission page.
Second, another thing that would be interesting to know is what the total page views are across the site (and including the off-site submission page). Pageviews matter in some cases as much as visits.
For example, consider these two sites:
SITE A – 3,000 visits a month, but the visitors only look at a page or two
SITE B – 1,500 visits a month, but the visitors look at 10 pages each on average.
In some cases – SITE B will be more valuable to an advertiser.
So, one thing PhilImprint could do is look to see how many page views they’re getting (in addition to visits). I might recommend Google Analytics to your web person. The stats collected by Analytics is amazingly comprehensive. I’m a big fan of it.
Third, related to the page view tallying – I noticed some things about PhilImprint that might hurt overall page views – but I’m not sure how the tracking/tallying works on your end so I can’t be sure. So these are some suggestions for maxing page views (but you may already have them implemented)
A. Have each page browsed in a journal article count as a page view. (I’m not sure if it does already…but if someone reads through the entire PDF of an article on your site in that neat PDF reader you all have…you want credit for each page view)
B. Make sure every page view counts toward philosophersimprint.org in your overall pageview tally. I noticed that some pages on the site don’t look like they’re part of the site domain at all. (e.g. the submissions page).
This is giving me some confidence that an open-access journal might be able to monetize with advertising, but it sort of depends on page views and how many visitors might not have been counted if they didn’t visit the site in the way you specified above.
I can see that page views, visits to the home page, the About page, and the Submissions page would be of interest to advertisers; but they’re not of interest to us as editors, or to the Scholarly Publishing Office as publishers. We’re very unlikely to collect these statistics. It would be much easier to start selling ads and see how many clicks they garner. (Not that we’re about to sell ads, mind you.)
As for our interest, as editors and publishers, in measuring our readership:
The page-view total is not meaningful. We offer two views of every article (as does JSTOR, for example): individual page-images and pdf. There is no way to track how many pages are viewed by someone who accesses the pdf. “Page views” can be tracked only in the page-image format. So they are a partial and not-very-meaningful statistic, when it comes to measuring our readership.
Our usage statistics can’t be compared to the circulation statistics of print journals, since the latter are just sales of the current issue: they don’t include retrieval of the journals in libraries or online. On the other hand we’ve been publishing for less than 10 years, and in many of those years we published only a handful of articles. So the literature contains many fewer citations to us than to journals that have been putting out four full issues a year for decades. We should therefore expect that people have many fewer occasions to retrieve our content, even if our individual articles have a larger than average readership.
One relevant (but impossible) comparison would be the average number of readers for an article published N years ago. Another relevant (but impossible) comparison would be the number of genuine browsers — that is, the number of readers who access the journal with the intention of reading around in it, which is not the same as the number who choose the “browse” view in order to locate a specific article.
Given all of the complications and limitations, I’m inclined to favor the policy of ignoring the statistics, all of which are fairly meaningless as measures of readership. Again, I realize that readership is not what would interest advertisers. But advertisers could count clicks rather than traffic on the site.
I agree that page views would be meaningless given the current way the journal is set up.
However, IF you could track each page of the PDF viewed that would probably be a better guide to readership than the current set of statistics. So I think editors should be interested in page views as a guide to readership (if the pages viewed in the PDF viewer could be tracked). There must be some way to do this. You readers are performing an action on your site to click through those PDF pages.
If this data could be obtained that would be very valuable. You would know not only how many visitors you had, but you would have some idea as to how much of the journal actually gets read by the visitors.
Imagine if Mind and Nous could both magically track how many pages had been looked at in their journals. Suppose they both have the same circulation. But suppose Mind has 50% more page views than Nous. It seems that Mind could boast better readership.
Of course my original thought that page views would be of interest to editors is conditioned on the ability to track the number of journal articles tracked. If that’s not possible, then I agree page views are probably irrelevant to generating meaningful data on readership.