MFA Immunity From eCPM Impact

Does Amount of Page Views Matter?

[Content_ed]

I'm always amazed when I read that some publishers have eCPM's as steady as a rock. Our eCPM mmoves with the number of MFAs that get through to advertise on our site, in extreme cases this summer eCPM can be half or double from one day to the next, as the MFAs come and go.

We've stopped worrying (for the main part) that we're doing something that encourages Google to cycle MFAs through our site, assumedly there are large numbers of arbs starting new campaigns every day and we just get our share.

What I'm wondering, is are the swings due to our being a relatively small publisher, with just a few thousand page views showing Adsense each day on our site? If Google has some fixed number of new ads they will dump on any given publisher in a day, rather than doing it based on traffic, it would explain large publishers not seeing any statistical difference while small publishers go nuts trying to filter the MFAs out.

[AndrewT]

Hi,

an interesting post, the sentiments of which I can relate to fully as I too see fairly big swings day in day out and, from looking at my site, its for the same reason you state. For that reason, my suspicion is that you are completely correct - but I'll be interested to see what others think.

I think the key is not to act in a reactionary manner - something I have done too many times with re-jigging the ads and so forth.

That said, it's a bad month all in all but having searched this forum and others, August seems to be a bad month for many (though by no means all).

In honestly, as much as I would like to feel in control of Adsense (and I read so much a day on the topic) I realize that it ultimately controls me (sorry, went off on a tangent there).

[jomaxx]

Smaller number of pageviews = bigger statistical swings, even if nothing whatsoever changes. I won't say MFAs can't drag down your revenue, but it's pretty hard to quantify the exact effect day in and day out. I would expect that a lot of the swings you see are actually just statistical noise.

[Content_ed]

I wouldn't be surprised if noise does play a part, I'm just not sure it's the large part. I can compare my management of Adsense on our site to last year and the year before. Previously, anytime our eCPM dipped below $XX.XX on a channel for a couple days, I would start checking the ads displayed. If I saw MFAs, directories, or poorly slected ads, I would add them to our filter (now full), and 100% of the time, the eCPM recovered the same day.

One of the reasons we aren't showing Adsense on 20,000 or more page views a day is that a great deal of experimentation showed that poorly converting pages didn't just drag down the eCPM, they dragged down the earnings for the whole site. I understand that this pruning of pages is still a matter of debate for webmasters, just reporting how it worked for us. We only display Adsense on a small fraction of our page views.

But the downside of running lean and mean seems to be sharp eCPM and revenue swings in the "off season". We are appreciably steadier outside the summer months.