Forum Moderators: martinibuster
My site isn't about just one thing, but many, and draws a wide diversity of ads across pages. If I'm depleting the stock of ads, I'm depleting multiple stocks.
Questions:
1. Does everyone see this?
2. Is this associated with topics with small stocks of ads? With low-priced ads?
3. What can I do to cut the rubber band? Should I "taking less" on weekdays—showing only one ad per page, not two or three—or would I just get a smaller slice of a smaller pie? My first (top) ad gets most of the clicks.
Anything else to do?
If a lower EPC is the culprit, it's possible that the drop is caused by factors that are beyond your control (e.g., too little inventory for high-traffic days, or bid managers who suspend campaigns over the weekends when they're at the beach or playing golf).
[webmasterworld.com...]
Obviously there appears to be a limited number of advertisers in some sectors and bearing in mind that the US is the second half of the day for many advertisers I can only assume that many budgets may have run out and that the US advertisers take advantage of the lower ad rates...in my widget industry anyway, thus resulting in lower EPCs even though you, and me, are getting more traffic and clicks.
This month is the first static month I have had since I started optimising my pages more efficiently for Adsense and I am now wondering whether I am actually hitting the ceiling for earnings within my industry?
For instance, I have a technology site that draws two types of visitors. Visitor A is a business man going through a midlife crisis who is interested in one of the products related to the technology. Visitor B is a student doing research on the technology. On weekends, when my traffic primarily Visitor A via SE's (for the product keyword) and directories, my EPC, CTR, and CPM for that site skyrocket. During the week when I get an influx of Visitor B's via SE's (this time for the technology keyword,) educational, resource, and research sites, my EPC, CTR, and CPM drop considerably.
For this particular situation, I found the best option to maximize earnings was to set a cookie for certain referrers that would show an ad (through CJ instead of AS) that earned a significantly higher CPM than the AS ads. By optimizing ads for both Visitor A and Visitor B, that site's CTR, EPC, and eCPM are now stable, and total earnings are much higher overall.
I've had a similar experience - with increasing traffic eCPM drops. I agree that are a lot of possible factors for this:
1. As you get more traffic, it may be less well targeted on average than when you had fewer visitors. Poorer targeting on some channels means worse conversions. As a result, you might be getting penalized through smart-pricing across your entire site.
2. Your high traffic could be depleting the high paying ads on one channel, leaving low eCPMs on the remaining channels.
and I'm sure others - but I'm tired of worrying about them.
My suggestion is to diversify your revenue streams. As traffic increases it becomes more worthwhile to move to CPA programs. With low traffic CPAs don't work so well - you send a few good visitors but are taking a big risk that they'll convert. CPC is better with low traffic. With higher traffic #s CPAs start making sense because you'll get enough conversions to make the risk worthwhile.
I used to live and die by Adsense - 95% of revenue came from it, with heart palpitations every time something changed on their system. In the last 6 months I've diversified, taking a 15% hit on Adsense revenue, but increasing revenue from CPA and direct advertising. Adsense is still the biggest earner (60%), but the other sources buffer the problems and inexplainable fluctuations in Adsense.
The higher my traffic, the lower my eCPM
I have the exact same situation, however mine is based on several sites and the combined eCPM of all the sites. Here is what happens to me.
One of my sites is a high traffic educational site with a very low eCPM. Most of the traffic on this site is from Mon-Fri.
My other sites have a consistent higher eCPM for all 7 days.
During the week my eCPM is lower, but on the weekend my eCPM is much higher without the traffic from the lower eCPM site.
You may be experiencing the same thing, but with different pages instead of sites.
I have 2 sites, similar themes, so displaying identical ads ( theres not that many ads in my sector ).
One site does about 100 impressions, the other does 2000.
A click on the 100 impression site is always worth 10 times teh click on the 2000 impression site.....
Getting a feeling am being taken for a ride here by google :(
I guess the way to handle it is to do lots of Excel work. So, I did a scatter plot of traffic against CTR, EPC and earnings. Both my CTR and my EPC go down slightly as the traffic goes up. Earnings obviously goes up, but the the trendline is a bent curve.
When I do only weekends or weekdays, CTR and EPC still show a downward slope as traffic increases, but not perhaps as downward.
To me, this means:
1. Traffic per se is hurting my CTR and EPC
2. All things being equal, weekends are just better
The question is: how does AS allocate ads? If you "take more" do you get worse ads, or is it merely about the "global" market.? So, for example, do my ads get bad because of my high traffic, or because my high traffic is correlated with AS-wide high-traffic?
The test here is to try to "take less," either one ad per page or stop displaying AS ads part-way through the day.
Another good test would be if all your guys took less, then I could take more. How about it? :)
How do they determine the income for an ad? I believe they look at income from your site, not across the network, though they must have to use some kind of averages on new sites, and then adjust when there is a history.
New ads, with no history, get spun into the mix and are allowed to sink or swim. The ones that get clicks stay in the rotation. The others gradually sink.
This system can have some awkward consequences, if you, for example, put two banner ads on a page, top and bottom, and your visitors are clicking mostly on the bottom ad, after reading the page. That means they miss the top-paying ads and are clicking the lower-paying ones.
I only use one ad block per page, mostly 4-ad skyscrapers. On some pages, primarily for aesthetic reasons, I use 2-ad blocks. EPC on those pages tends to be higher, I guess because visitors are only seeing the two top-paying ads. CTR tends to be lower, so I'm not sure I want to go to the 2-ad blocks site-wide.
[Added to answer about allocating ads per site] I see the same ads appearing on page after page on my site, all targeting the same or similar keywords. There is no limit to the number of times a particular ad appears on a site. It will simply appear on the pages where it "works." If you add pages on slightly different topics to an existing site, they will get different ads, perhaps with some ads that target the overall theme of the site mixed in. Or at least that's what I see.
The traffic source and entry pages are easier to see, so that's where I'd start to look for reasons. The entry page a listed in a se refer might give some clue to the motivation behind the visit, but I doubt that'd be very reliable, butit might still worth looking at if possible.
Anther thing might be the time of the visit vs the type of page content involved.
People visiting during working hours may not be able to spend the time to explore much. But they might return after work or on weekends and be more inclined to dig around in the site more, spend the time to make a purchase, or click on ads.
That sort of behavior might account for some of what you are seeing.
In my experience, visitors who limit their page views to my photo galleries are apparently a lot less likely to click an ad.
But if they click through to one of the other sections of the site the likehood they'll click an ad goes up noticably.
The result seems to be that when entry traffic drops off in my photo galleries my stats seem to improve.
I'm sure I haven't figured this all out yet though, so I could be way off the mark here.
But I seemed to draw some kind of targeting penalty as a result. The site that drew the low eCPM that day seemed to drag down my others. I hoped in vain for a bounce, but it was slow in arriving. So I pulled ads entirely from that site. This did wonders for my overall CTR. I've commented previously about some sort of breakpoint at 2% CTR. Like others have previously reported, I'm seeing a boost from dropping ads on the worst-performing pages in my network.
I think in my case there are a mix of reasons for the site to perform poorly, and that's not so much my point here. The original question posed, "What can I do to cut the rubber band?" is ineresting, and I propose that one or less ad per page may cause you to see a net improvement. It did for me.
I see eCPM bounce around from day to day, sometimes in a discernable pattern, mostly down, thn a good day, up, but lately it's been up, down, up, down, and no single day or weekends vs. weekdays stands out. In fact, my best earnings and my worst earnings both came on a Friday.
Fiddling around with one's site in order to increase or decrease a single or set of variables, is, to me, counterproductive to knowledge and is an overall detriment to the purpose of publishing.
In other words, trying to max out adsense earnings, especially when we're working in the dark, is a massive waste of time and effort.
I'm sure some of you may disagree with this opinion, but I think I have learned enough about ad placement from 20 years in the newspaper and grphic arts business to know where and how to display ads. The Google fluctuations are more of an annoyance to me than my time is worth.
If G can't determine reasonable values for publishers, I'm quite certain that the market will take care of that problem in due course.
Fiddling around with one's site in order to increase or decrease a single or set of variables, is, to me, counterproductive to knowledge and is an overall detriment to the purpose of publishing.In other words, trying to max out adsense earnings, especially when we're working in the dark, is a massive waste of time and effort.
The non-linear relationship between an increase in traffic and income feels real to me--feels like a glass ceiling.
Given the number of bright minds on this board and the number of times people have tried to manage "the bounce", I agree with fearlessrick, we seem to be guessing in the dark and wasting our time. More information would be helpful, but those details would distract us from "creating more great content".
OK, everybody back to work! :)
Wise words for sure, and as EFV always reminds people, build it with lasting appeal.
it is so tempting to attribute increases or decreases in revenue or eCPM to the tweak you just made, but if google has introduced a small delay to the equation, everyone gets thrown for a loop.
us adsense publishers monitor things so closesly, that a delay of just a couple of days or a week, could be really confusing.
by the way, i think they intentionally introduce a much longer delay on the SERPs to really confuse SEOs.
I went and removed Adsense from about half my website and...
... my overall revenue went up. I'm earning more from Adsense by displaying fewer ads.
I feel like I'm in some kind of Bizarro world.
Think about two theoretical visitors:
Visitor 1) hits your page looking for a particular product or service, sees an ad for it and clicks. He is gone after 1 page impression.
Visitor 2) hits your page, surfs around 10 pages at your site learns about the product or sevice and finally clicks on the same ad as visitor 1.
Both visitors earn you the same amount of money but visitor 2 only generated 10% of the eCPM as visitor 1.
By careful tracking you can easily follow this type of visitor behavior.
