| 10:13 pm on Jun 4, 2006 (gmt 0)|
> I'll spend the next two hours cleaning them up!
I've been flip-flopping on what to do about bad ads for sometime now. Recently I started using the competitive filter to block over twenty of the worst offenders. However, I'm now down to less than five - all bogus survey ads. Much as I don't like the idea of the big auction sites/portals and MFA lame ads appearing on my sites, I really feel like it's over my head trying to police them too. There's simply too much statistical info that Google is not affording me to be able to know what using the competitive filter to guard my sites is really doing.
| 10:36 pm on Jun 4, 2006 (gmt 0)|
Try my EPC-based strategy [webmasterworld.com] as a more permanent solution to MFAs. It'll probably increase your earnings in the process. Basically, the strategy involves managing your ad stock to increase the value of each ad space and price MFAs off your site.
| 11:05 pm on Jun 4, 2006 (gmt 0)|
Hmmm. I must have missed this post back in April.
I'll check it out 21. Thanks.
| 11:07 pm on Jun 4, 2006 (gmt 0)|
About twenty four hours ago I had noticed a very large decrease in ecpm and in my earnings. I spent the greater part of the following hours culling MFA's that were displaying on my site across various countries. By the close of the day yesterday, my ecpm had increased substantially
Now my ecpm has bounced back up with a vengence as it has doubled on yesterdays's starting figure. It's actually back to Feb-March levels at the moment.
Quite surprisingly my ctr is also much higher and back to Feb-March levels also. Earnings have increased but remain lower than FEB-March figures; however that has more to do the decreased traffic rather than the lower average CPC that I have been experiencing lately.
| 11:42 pm on Jun 4, 2006 (gmt 0)|
|Try my EPC-based strategy as a more permanent solution to MFAs. It'll probably increase your earnings in the process. Basically, the strategy involves managing your ad stock to increase the value of each ad space and price MFAs off your site. |
That's exactly how I've been managing our site lately, even dropping some OK pages to force the EPC higher. Seems to be working so far, making about the same money with around 80% of our pageviews Adsense free.
| 1:37 am on Jun 5, 2006 (gmt 0)|
Good three, 21. I do recall reading that back in April, but I may have been experimenting with something else at the time.
Either way, I have been aggressively removing ads from pages with very low EPC for the past three weeks and my EPC has increased dramatically. I've gone from $0.05 (lowest I've seen) to $0.41 (average).
Targeting has also been a lot better, which is why I was shocked to see MFA's taking over today.
| 3:49 am on Jun 5, 2006 (gmt 0)|
Just blocked another 20 or so MFA's. The more I block the more shows up.
Anyone seen one whose landing page is a google search results? Had 2 of them.
| 3:53 am on Jun 5, 2006 (gmt 0)|
|Just blocked another 20 or so MFA's. |
Same here. They are so many and widespread that it's beyond a joke now. They make my product site look like a cheap spam scam site.
Once again my filter is almost full
| 4:12 am on Jun 5, 2006 (gmt 0)|
Yeah, same here.
| 9:40 am on Jun 5, 2006 (gmt 0)|
celgins, another thing to look at is whether there are some poison words [webmasterworld.com] that are letting MFAs into your site (and the technique in the current thread about negative keywords may help identify them). Or it could be poor performing pages [webmasterworld.com] (msg 10).
Imho, the basic key to keeping MFAs off your site is making sure that high EPC ads individually have a healthy CTR - also discussed in April [webmasterworld.com].
The CTRs of good ads don't have to be as high as MFAs, but high enough to make good ads' eCPM higher. In the first instance this means reducing the number of ads on a page, not removing them entirely. Eg: if you have 4 ad slots on a page (eg in a leaderboard) with a CTR of 10%, that means each individual ads' CTR is, on average, 2.5% (10 divided by 4). But if you reduce the no of ads to 2 (eg in a banner), even if overall CTR for the page drops to 8%, the individual ads' CTR rises to an average of 4% (8 divided by 2).
So reducing the no of ads on a page can improve good EPC ads' eCPM on your site, making it higher than MFAs (which have high CTR and low EPC). MFAs are also unable to creep onto the bottom of pages because there is now much higher competition for fewer spaces.
| 1:28 pm on Jun 5, 2006 (gmt 0)|
|In the first instance this means reducing the number of ads on a page, not removing them entirely. |
Are we the only ones experimenting with lowering the overall number of pages with Adsense? We stuck with no more than one ad unit per page pretty much from the beginning, and because we're more concerned with selling widgets than ads, we never have the Ads displayed above the fold. After a while of monitoring channels, it became apparent that Adsense was following life's 80/20 rule, around 80% of our revenue was coming from 20% of our page views. Normally, we would have been happy to keep collecting that extra 20%, but the the targetting and MFA problems started, and overall revenues began dropping. For the six months from last September to this March, we ended up with Adsense on maybe 25% or 30% of our page views, and earnings seemed pretty steady. Then the MFA's returned in a big way and we moved to heavy use of the competitive ad filter. Finally, we tried Adlinks a couple month ago to try to get more page views going, but that really killed eCPM, brought more MFA's, worse targetting, etc.
Currently, we're all the way down around 20% of our page views carrying Adsense, eCPM is at an all time high, and MFA's seem to be priced out. While earnings are trending around 80% of "normal", if that keeps the MFA's away and holds revenue steady, it's a good deal. Besides, we weren't running Adsense at this level last summer, so I don't know if some of the fall off isn't just seasonal, our widget sales are probably down 20% to 30% every summer in any case.
| 1:52 pm on Jun 5, 2006 (gmt 0)|
|Are we the only ones experimenting with lowering the overall number of pages with Adsense? |
No, you aren't. In fact, I think your approach is the more common. Eg: there are lots of threads in this forum about removing ads from poor performing channels (which mostly means pages).
However, the techniques in the links I've posted are designed to get a lot more income from the tail. For example, on one "low performing channel" I've managed to increase the income by 250% since last October by cutting down on the number of ads displayed. I think a lot of other publishers would have removed the ads altogether. But that loses out on potential income - OK, not a great amount, but it adds up if you have a lot of pages like that.
There are pages where ads have to be removed altogether. But for our site that is a very small number - although we have a core of good earning pages, we have managed to get a lot of other pages performing moderately well, which keeps our overall income at a high level.
| 2:08 pm on Jun 5, 2006 (gmt 0)|
|No, you aren't. In fact, I think your approach is the more common. Eg: there are lots of threads in this forum about removing ads from poor performing channels (which mostly means pages). |
My bad, I should have said "in this thread."
I do wonder how just reducing ads works so well for you, when we never had more than one ad unit per page to start with but saw the same dilution effects. Maybe it has something to do with overall number of page views or the subject areas we're in.
| 2:22 pm on Jun 5, 2006 (gmt 0)|
|In the first instance this means reducing the number of ads on a page, not removing them entirely. |
Sure. I've been doing this as well. Unpopular pages with very low EPC are removed altogether, while pages with two or more ad blocks are reduced to only one ad. When reducing a page to only one ad, the targeting is 10x better.
What's so strange is the way Google promotes better success with AS. In a few places within their documentation, they mention adding more ads to more pages to help increase earnings. I think this is why so many AS newbies start in SmartPrice Hades!
| 2:25 pm on Jun 5, 2006 (gmt 0)|
Lowest ever earning in more than a year this Sunday. i do not monitor ads on Sunday so I donít know if I am done with MFAs
| 2:29 pm on Jun 5, 2006 (gmt 0)|
|we never had more than one ad unit per page |
Even with one ad unit you still have control over the number of ad slots that can appear, from a single ad (eg: a button) to five (eg: a wide skyscraper). And the latter dilutes the CTR of good ads by a factor of 5, which can perhaps bring the price (ie: the eCPM associated with the ad slot) within range of an MFA.
So, when I'm talking about reducing the number of ads, I'm not counting ad units, I'm counting ad slots. That is:
wide skyscraper = 5 ad slots
leaderboard = 4
skyscraper = 4
medium/large rectangle = 4
square = 3
banner = 2
vertical banner = 2
half banner = 1
button = 1
small rectangle = 1
| 2:35 pm on Jun 5, 2006 (gmt 0)|
|Even with one ad unit you still have control over the number of ad slots that can appear, from a single ad (eg: a button) to five (eg: a wide skyscraper). And the latter dilutes the CTR of good ads by a factor of 5, which can perhaps bring the price (ie: the eCPM associated with the ad slot) within range of an MFA. |
That's a great idea, if we see any new signs of erosion I'll give it a shot. I guess I was lulled into the notion that Adsense would optimize the number of ads shown in the single block, which I do see changing from time to time, but we all know that their algo is far from perfect.
| 2:39 pm on Jun 5, 2006 (gmt 0)|
I'm noticing that the ad inventory seems to vary quite a bit for certain days. Specifically, I think some AdWords advertisers are only advertising during the week (and mainly during prime business hours).
My proof? One of my sites it dedicated to smartphones (PDA + phone combination devices) -- this weekend it was absolutely overrun with ringtone ads. Which, of course, don't pay squat. Every single ad block was ringtones at one point. This morning? Nothing but PDA's, cell carrier service plan specials, etc. -- the type of targeting that I would expect (and want) on my site as it pays much better.
What did I do between yesterday and today? Nothing. So, as the ad inventory changes - you'll notice better ads (more inventory, better targeting) or worse ads (very few advertisers paying, which means the MFA's and other junk starts getting placement).
So, I wouldn't stress too much about a momentary spike in the number of MFA's or junk ads you see - especially when that spike comes over the weekend. I suspect that a lot of the premium AdWords advertisers stop their ad runs outside of prime hours.
| 2:44 pm on Jun 5, 2006 (gmt 0)|
|if we see any new signs of erosion I'll give it a shot. |
There can be a problem in that, once the erosion starts (ie the MFAs start to appear) then reducing the number of ads can simply result in the MFAs becoming the only ones that appear.
As David_UK said in one of the above threads, it is like trying to control a disease, so think SARS or foot-and-mouth: keeping it out is easier than dealing with it once it is in. The way I preempt MFAs is to try to keep the average EPC at a high level (in relation to the Adwords price). This tends to result in having a page with few ad slots, above the fold.
Having said that, there are some pages where I have 8 or more ad slots throughout the page, and I get very good returns. I reckon there must be a lot of advertisers in that niche.
| 4:26 pm on Jun 5, 2006 (gmt 0)|
Here's a snippet [webmasterworld.com] from 21's post in that previous thread he mentioned.
|MFAs spread from the lowest performing pages upwards; they become the dominant ads because of the combined impact of smartprice and the MFA's high CTR sucking poor performing pages into their reach, a bit like a black hole gobbling up the universe (apologies if I'm using too many metaphors!) |
I think this is one of the most important notices AS newbies should be aware of. I was one who followed AS guidelines and placed more ads on more pages.
As a result, EPC dropped and my earnings took a dive. I run a magazine site, so there are dozens of pages with high popularity; while others are very low.
Should not have placed AS on those low popularity pages.
| 7:02 pm on Jun 5, 2006 (gmt 0)|
celgins, just to ensure clarity, I was talking about page "performance" not "popularity". And for most of my techniques I define performance in a very specific way (by comparing average EPC with the going rates in Adwords).
Low popularity pages can be high earners. For example, I have one page that is responsible for less than 0.5% (half a percent) of my page views, yet produces 4% of my income. Another page has less than 0.15% of page views, yet produces 3.2% of income.
In fact, I have removed ads completely from my most popular page (7.5% of pageviews) because, apart from one month in my early days of adsense, it makes hardly any money and the EPC values suggest it has subsequently had a strong downwards impact on smartprice.