Fun with CPM ads (not)
by Sapphire (December 7, 2007)
Please note the response from BlogHer’s rep in the comments.
Recently, one of my sites experienced something like the Digg effect when a single link on a popular site sent hoards of traffic. Unfortunately the hoards were mostly trolls, not genuine new visitors. But this should have been a good thing all the same, thanks to CPM.
And my BlogHer ads made about $100 in one day. I was impressed!
Then I started talking to a friend who’s looking into CPM for her sites. She found a calculator on BlogHer’s site and punched in her estimated pageviews for a month - 300,000. The calculator said she would earn $1500 for that. She was unimpressed, given how it breaks down per pageview. I was very impressed since I’m not making anywhere near that.
Then I checked the stats on my site that had the traffic spike, and realized I had over 341,000 pageviews last month. WTF? Altogether, my BlogHer CPM earnings were just over $200 instead of the $1700+ their calculator predicts. I looked more closely at my earnings report:
- BlogHer reports more like 185,000 pageviews. Now, AwStats and any other stat package can come up with radically different numbers - it’s no surprise a CPM advertising network would use a more conservative stat package. And they’re geo-targeted. But that’s not all!
- BlogHer’s “default” ads - the ones you don’t get paid for - numbered around 125,000. That means I only had a 1/3 chance of getting paid for the pageviews BlogHer recognized.
- I also had several thousand non-paying hits for charity campaigns.
- A lot of the campaigns right now are, like, $4/1000 hits rather than the usual $10 or $12 because they don’t have enough campaigns at the moment.
Now, BlogHer is fairly new and it’s possible things will improve. But from the research I did yesterday, this doesn’t sound too abnormal. At least not among the sort of networks I stand the slightest chance of getting my sites into. BlueLithium, for example, does the “default” ad thing, too. Plus I found this in their ToS:
Publisher understands that any violation of this Agreement will result in forfeit of any money owed by Blue Lithium to the Publisher as well as legal action against the site. In addition, as a result of the violation of this agreement by the Publisher the Publisher will pay to Blue Lithium, within 30 days of being notified of such violation by Blue Lithium an amount of $30,000.00 (US), plus any legal expense as well as damages incurred by BlueLithium.
And now we know why “lithium” is part of their name. Somebody get them some! Stat!
I’ve applied to several other CPM ad networks (ValueClick, Casale, a couple of others… and oh, yeah: Federated Media, just for laughs while I don’t hold my breath). If any of them will accept any of my sites, I’ll let you know how they perform. If it’s not a lot better than BlogHer, then I’m afraid CPM may just not be a great solution.
Posted in Ad Networks
Subscribe to the feed or email updates
December 7th, 2007 at 9:55 am
Hello! I wanted to explain the ‘default’ impressions that occur in our reporting.
Our ad tags are written to allow up to three ads to appear in your sidebar. Since we typically have only one ad at a time - the other two tags remain blank. Each time your page loads, the tags ping the tracking server - once for the ad that is display, and two blank or “default” pings.
As you noted above - these default pings typically represent about 2/3 of the reported impressions - but do not represent additional page loads. (If you compare the non-default impressions with your other stats tracking, it should be close.)
One other quick note - the house ads, at $4 CPM are the equivalent of an $8 CPM, since BlogHer is not taking a share of this revenue.
As always, we’re here to help if you’ve got futher questions or concerns about BlogHerAds.
Best regards - Jenny
December 7th, 2007 at 1:15 pm
I hope it was clear, Jenny, that my frustration is with CPM in general, not BlogHer specifically. It’s just BlogHer’s the example I have on hand. I think you guys have been fair, and the fact that you don’t take a cut of the house ads is wonderful. Also, you have those headlines that bring us some traffic from within our niches - I think that’s brilliant.
However.
If I’m adding correctly, your system considered only about 50k of my 340+ AwStats pageviews to be real pageviews. That’s fine, I get the reasons for that. But AwStats is sort of industry standard for webmasters, so what I have to do to evaluate BlogHer alongside other monetizing solutions is translate things into AwStats terms, and that means I need over a million AwStats pageviews to get about $600/month.
That is a lot more traffic than most indie webmasters are going to get. Therefore, unless other CPM networks can deliver, say 2-3 times more revenue to a site like mine than BlogHer, I’m seeing the limits of what CPM can bring me. And in case it wasn’t clear, I’m skeptical as to whether anyone can do THAT much better than BlogHer. I have a feeling maybe CPM just isn’t the ultimate solution for some sites.
December 7th, 2007 at 6:05 pm
Hi Sapphire, interesting post. It made me want to chime in!
Sorry to hear you are disappointed with your CPM revenue earnings. But I wanted to address a more interesting issue that I see in your post - and that’s with AWstats. Yes, it is a standard free web analytics package that comes with all cPanel servers. No doubt that it’s pretty widely used out there. But, AWstats, in my own experiences, is not that accurate when it comes to pageviews. In fact, there is no such thing as pageviews in AWstats, only something called pages.
Go to Navigation >> View >> Full List and you should the pages that AWstats measure. In one example, I can see some php and cgi scripts being included as a separate page that was viewed. It may be correct that it’s a file that was called but it’s not a true user initiated page view. Calling these files will never generate an ad impression. Only a full pageload will yield an ad impression. So, even if you have a few elements that make up a page load, you should still only receive 1 page view. With AWstats now, I see it counts these incidents separately, thus falsely inflating pages viewed. Your stats could be mis-reported like this as well. As a marketer, I don’t really need to know how many times scripts and files were called. It’s much more important to know how many real pages were viewed on a site.
Based on my own experience, the free Google Analytics (GA) is a much better indicator of true pageviews. I have compared GA and AWstats with multiple networks (CPM & otherwise). GA definitely gives a much more accurate pageviews (as compared to totals ads served by CPM networks). Just copy and paste 1 simple tag into your site template to install GA. It’s free by the way: http://www.google.com/analytics/
I noticed you refer to Quantcast on your advertiser page. You should get quantified quantified by http://www.quantcast.com so that your potential advertisers will see a more accurate result. Quantcast is not known for being too accurate, especially with blogs. Getting quantified will give you a decently accurate result for pageviews and visitors as well. This one is also free.
You also addressed fill-rate from CPM networks. I will have to address that in another book. Sorry for the lengthy post. Hope it helped and good luck with your revenue generation!
December 7th, 2007 at 6:28 pm
I absolutely won’t use Google’s analytics, or Google anything. They already own the web; I’m not just handing them data on a silver platter.
I did have Quantcast set up. It says I had between 150-200k pageviews during my traffic spike. Still not really matching up with BlogHer’s 50k.
But again, it’s not the stats accuracy I’m concerned with. All stats are “accurate” - it’s just that they use different methods to calculate what’s a pageview, what’s a real visit, how long someone has to go away before returning to be considered another “unique” visitor, etc. I have no problem with BlogHer using whatever they use.
I do wonder if my host cached pages or something that could have affected the count, though. Could be a lot of things.
December 7th, 2007 at 11:32 pm
So it sounds like your alternative is to get accepted into other CPM networks, run their ads and measure ad impressions. Remember, most of the time, pageviews is almost always 100% > ad impressions (up to a 20% discrepancy when pageviews are high).
As for “All stats are “accurate””, I look at it another way. Accuracy is not the same as precision. A weight measuring scale can be very precise but it may not be accurate at all. Ahhh, brings back memories from the good old stats days.
December 8th, 2007 at 11:44 am
found your site thru Ec–i like it, nd have subscribed.
December 11th, 2007 at 11:35 am
[...] 11, 2007 One of my sites - the same one that experienced a Digg-like effect - crashed (or helped to crash) the DownTownHost shared server it’s on yesterday. There was no [...]
February 14th, 2008 at 12:23 pm
[...] few months ago, I wrote about my BlogHer earnings, and how far off the pageviews according to AwStats were from what BlogHer calculates, which resulted in around $200 of actual earnings for over 341k pageviews (according to AwStats). [...]