| This 47 message thread spans 2 pages: < < 47 ( 1  ) || |
|Declining eCPM - action checklist|
What to do when your eCPM continues to drop
| 9:55 am on Jan 25, 2006 (gmt 0)|
I've noticed a slow but steady decline in the eCPM associated with my account over the past 12 months. Compared to January last year traffic is up four-fold but eCPM is down by about 50% (can I say that without breaking the TOS as long as I don't mention specifics figures?)
I'm a fairly regular reader of this forum and I know that a lot of people have seen the same thing (traffic up, eCPM down) but there are also those who see the reverse, or no change at all. I appreciate there are some things out of our control such as the number of advertisers and the amount they're willing to bid at any one time, but leaving that out of the equation are there steps we can take to give us the best chance of at least maintaining a good eCPM?
So far, I've tried the following:
- Removed underperforming ads
- Reduced the number of ads per page
- Experimented with different colour schemes to maximise CTR
Unfortunately I've met with little success. Are there other steps I can/should take? For example:
- Does traffic from the AdWords content (rather than search) network have the potential to negatively impact AdSense eCPM due to a lower conversion rate, or are content related clicks one of those figures not factored in to the AdSense eCPM?
- Is there any point in contacting Google to ask if smartpricing is in effect (I'm guessing they're unlikely to give an answer to that, at least not a specific one as far as individual sites within an account are concerned)?
Perhaps we could develop a list of actions to take for anyone in the same boat, any suggestions gratefully received! If there's already such a list and I've missed it please forgive me and point me in the right direction ;-)
| 11:57 am on Jan 27, 2006 (gmt 0)|
my CTR seems do be dropping slowly, month by month as well.
| 2:15 pm on Jan 27, 2006 (gmt 0)|
> Yesterday (26.01.2005) was the lowest point in earnings since october...
> The traffic is normal, payment is dropping
Yes I noticed AdSense earnings dropped by 50% yesterday, continuing today. Low click through rate and low earnings per ad. I figure it is just a Google funny and will wait a bit to see what happens.
So far Chitika has not been worthwhile for me.
| 5:04 pm on Jan 27, 2006 (gmt 0)|
I see an ebb and flow of CPM based on which advertisers are buying at a given moment. I find it useful to rotate a variety of ads, including Adsense and YPN - if Adsense is tanking and YPN is running a better eCPM (or vice versa), I can keep cranking up the better program. Of course, it may not be completely scalable, i.e., the eCPM at 10K exposures per day probably won't stay constant if you ramp the program up to 400K exposures.
On low volume sites, I don't split the ads but periodically test an alternative network for a few days.
| 5:12 pm on Jan 27, 2006 (gmt 0)|
my average eCPM for the past year has been $53.19, how does that compare to everyone else`s eCPM?
My CTR has been around 13% for the past year as well.
| 11:01 pm on Jan 27, 2006 (gmt 0)|
|Sorry, palomar, but your understanding of how eCPM is calculated is incorrect. The Adsense Glossary states "It is calculated by dividing total earnings by the number of impressions in thousands. --21_blue" |
Yes your are corect, that is Google's definition but that is not the only way to calculate eCPM. Don't forget that CTR = clicks divided by impressions; it's just another way of saying the same thing. (4*3 is the same as 2*2*3)
My original point was to show that eCPM does not always go down when impressions go up, as was orignally claimed. That can be verified by the examples I gave.
| 12:43 am on Jan 28, 2006 (gmt 0)|
|My original point was to show that eCPM does not always go down when impressions go up, as was orignally claimed. That can be verified by the examples I gave. |
I agree with you that "eCPM does not always go down when impressions go up", a statement that is fully compatible with what I'm saying.
From one point of view the difference between what we are saying could be dismissed as just a question of semantics. But I think it is important to clarify because it seems that some members of the forum have a misplaced faith in and over-reliance on eCPM. Whilst eCPM is useful for some measurements, it is not useful in all circumstances, and using eCPM to support decision-making 'in all circumstances' may result in a loss of earnings.
To try and clarify: I didn't claim that "eCPM always goes down when impressions go up". What I said was that, by definition, if impressions go up then eCPM goes down. However, in practice eCPM could still increase as impressions increase, if and only if earnings (the other half of the equation) increase by a greater proportion than impressions.
My point is that if you increase impressions alone, eg: by you or your ISP speeding up your site, or improving site stickiness through a better navigation system, then by definition eCPM and CTR will go down.
In fact, our experience is that in those circumstances, earnings also go up. Ie: we've made changes to our site; CTR has dropped; eCPM has dropped; but earnings (per visitor) have increased. If we had used eCPM to measure the effectiveness of those changes, we would have concluded (incorrectly) that they were bad changes, reinstated the original pages, and we'd have lost money as a result.
| 8:02 am on Jan 28, 2006 (gmt 0)|
| 8:40 am on Jan 28, 2006 (gmt 0)|
Switching underperforming pages to YPN helped me increase eCPM considerably.
My eCPM in jan. 2006 went up by 120% compared to jan. 2005
on my sites it seems at least that low performing pages really drag down the rest, once i got rid of them the rest performed (eCPM wise) much much better...
| 10:04 am on Jan 28, 2006 (gmt 0)|
|your ISP speeding up your site |
Lets call that Hosting 21_blue to avoid confusion.
| 10:07 am on Jan 28, 2006 (gmt 0)|
Thanks very much for that, I'll take a look right now!
| 1:17 pm on Jan 28, 2006 (gmt 0)|
|Lets call that Hosting 21_blue to avoid confusion. |
With what can 'ISP' be confused?
| 5:40 pm on Jan 28, 2006 (gmt 0)|
|it seems that some members of the forum have a misplaced faith in and over-reliance on eCPM. Whilst eCPM is useful for some measurements, it is not useful in all circumstances |
I don't think anyone has ever suggested that is. Obviously, total revenues also need to be taken into account.
The reason eCPM is used by publishers (i.e., media publishers, not just AdSense "publishers") is that it lets the publisher compare different revenue sources.
eCPM is also useful in comparing performance for different areas of a site. Let's say you've got a travel site about Sweat Island, Florida with travel-planning articles and photo galleries. You're getting eCPMs of $10 and $1 from the two sections (since the articles get decent clickthough rates but the photo galleries don't). If you know you can get a higher eCPM than $1 from a display-ad network such as FastClick or Tribal Fusion, you might decide to run display banners or skyscrapers in place of AdSense units on those low-eCPM photo-gallery pages.
| 6:25 pm on Jan 28, 2006 (gmt 0)|
|I don't think anyone has ever suggested that is. Obviously, total revenues also need to be taken into account. |
EFV, I think we are in agreement. I agree with you that eCPM is useful in the situations you have described. My comment about "misplaced faith in eCPM" was directed at the line of thought, oft displayed in this forum, that celebrates eCPM going up when Adsense is removed from low performing pages. Such thinking can be flawed, because it doesn't take total revenue into account.
Not for your benefit, EFV (because we agree that total revenues need to be taken into account) but for others, here are some fictitious stats that illustrate the point.
Suppose a site consists of 2 pages, each with 1,000 impressions and the following stats:
This gives an overall site CTR of 5.5% and eCPM of $12.
- CTR of 10% and eCPM of $20.
- CTR of 1% and eCPM of $4
If you then remove page 2, because it is a "low performing page", and nothing else changes, then the following happens:
This increase in eCPM might seem to be a point of celebration. However, overall earnings drop by $4.
- overall CTR jumps up from 5.5% to 10%
- overall eCPM jumps up from $12 to $20
If one counter-argues that there might be a compensating impact on smartprice, then maybe, but maybe not. In this example, EPC on page 1 is 20c and page 2 is 40c. If they are both about the same topic, this might suggest that the second page converts more. So by removing the second page, you are also removing the best-converting clicks, and the site smartprice is going to drop, which will hit earnings even more. The resulting eCPM may still be higher than your original figures, but the number that really matters - earnings - will have dropped.
| 7:30 pm on Jan 28, 2006 (gmt 0)|
21_blue, thanks for explaining things so much more clearly and succinctly than I could. :-)
| 8:25 pm on Jan 28, 2006 (gmt 0)|
My comment about "misplaced faith in eCPM" was directed at the line of thought, oft displayed in this forum, that celebrates eCPM going up when Adsense is removed from low performing pages. Such thinking can be flawed, because it doesn't take total revenue into account.
I should have added that total earnings as well eCPM went up considerably when removing low performing Adsense and replacing them with YPN...
| 11:06 pm on Jan 28, 2006 (gmt 0)|
|I should have added that total earnings as well eCPM went up considerably when removing low performing Adsense and replacing them with YPN... |
Sure, and in some cases your earnings could go up if you replaced decently-peforming AdSense ads with something else. (I'm currently testing AdSense-free pages on several sections of my Web site to see what impact the change might have on affiliate bookings, CPM ad impressions from display ads, and total eCPM.)
| 1:46 am on Jan 29, 2006 (gmt 0)|
>>>Now when things on AdSense take a dive, I shift 100% of my inventory over to Yahoo. I can't think of a better way to let them know how I feel.<<<
that's exactly where i'm headed.
i have a well documented situation where a much smaller site with a much better ctr and ecpm ends up bringing the overall earnings down drastically on both sites... it's a situation where i can put up viral content on the small site that people jump on the minute that it's posted, which drastically spikes the overall traffic.
so one day i gave adsense the most clicks ever, half of which came off of the small site, and the next day i got penalized heavily across all sites... both ctr and ecpm dropped drastically, to record low levels, on what was normally the biggest day of the week... and the last half of this month is the worst that it's ever been.
the subject matter is very similar, so both sites get the same visitors and the same advertisers... i am handing adsense fully qualified, relevant clicks via totally original content, stuff that took me years to create by hand... 100% white hat, no mfa garbage, and the adsense that people click on are the most relevant ads that they can possibly be, because i keep the url filter full.
my biggest regret right now is that i didn't start out with some kind of an ad rotator program in place... it's all hard-coded adsense right now, but that will be changing very quickly... i would encourage anyone reading this to not make that mistake... plan on rotating your advertisers from the start, you'll be much happier in the long run.
the day is rapidly approaching when adsense won't be the only game in town.
| This 47 message thread spans 2 pages: < < 47 ( 1  ) |