Forum Moderators: martinibuster
Wow, shocking! For how many days has it been running at 50%?
Not saying that it is due to the change, but seems rather unusual.
Started to see an increase late yesterday evening (usually the amount goes down as the day goes by, yesterday started to increase in the stats about 9:00pm PST)
Nikke, I'm curious. What's a page full of stop words?
Thanks for this Nikke, I'm an adsense newbie just decided to give it a go. Because I'm in books there are loads with death or murder etc in the title :(
Wondered what was happening on some pages
Off to try and find a list, if there is one.
Thanks again
Rod
Have some people seen the Earnings per Click thru drop or just me?
ECPM is down about 11% and CTR is down 0.3%
Not good, takes hours to block them too.
Taken from an affiliate support page:
Many of you are aware now that Google and Overture are getting alot more strict with affiliates. Because of this, we have decided to help you by providing you some new templates of the existing sites and show you through this document how to purchase a domain name, a web hosting plan and hide your site within a frame. Google and Overture will not allow two advertisers to have the same displayed URL and site. In addition, they will start verifying the ownership of the sites. This means that each individual advertising through search engines will have to get their own domains.
Google we need better filtering tools! [webmasterworld.com] - more filters allowed and perhaps regex-like conditions.
[edited by: amznVibe at 7:59 pm (utc) on Jan. 14, 2005]
I am seriously wishing this is delayed reporting but a pit in my stomache is telling me otherwise.
If it falls any lower, it may be time to dust of the Burst and Fastclick code (it is getting that bad in some areas!)
Moral of the story: diversify!
Filtering is taking ages and is soon going to run out of space, some affiliates have dozen of domains circumnavigating googles rules.
One question comes to mind.
Is it filtering out the same affiliate-advertiser a bad idea considering that the ads shown are the top paying ads? We could be shooting ourselves in the foot by ending up showing the lower ppc ads.
And yes ,revenue is down considerably
What do you say>
Filtering is taking ages and is soon going to run out of space, some affiliates have dozen of domains circumnavigating googles rules.
Moreover they seem to have bid on about every word possible so they are getting to all of our sites. I got one from a key/wordphrase like "wedding flower widgets". That was the title on my article and that was the title of the ad. Do they have some system where they can have the ad change to the page or article title? It makes for really misleading ads because it looks like it is just what the reader is looking for.
A couple of these have been for some polling site that has absolutely nothing to do with the topic on the page or the matching title of the ad.
People here have mentioned that maybe we shouldn't filter them as they are the higher paying ads over others we might get. I want to point out that an ad is worthless if it isn't clicked on. If the ad leads to something that our visitors aren't interested, and worse yet they feel they have been tricked into clicking, people will soon treat text ads like they do banners now and just ignore them.
Also don't forget most site visitors still think the people running the website are picking the ads. The sites reputation is at stake.
I am tempted to remove all the filtered names for 24 hours just to see what results I get. Lets see what happens in the next 48 hours.
If we can filter out the bad behaving affiliates with all the domains easily, the market will self-regulate itself. It simply won't be profitable for affiliates to behave like that so they will have to stop.
Since we can't stop them now, there is no end in sight to this problem - don't expect Google to be able/willing to check hundreds of thousands of affiliate domains without some well scripted automation that doesn't exist (yet).
I think this is a misconception. I could be wrong, but I think the terminology they use is "best" ads not
"highest paying". 'Best' to G is PROBABLY based on total CPM (a factor of CTR times bid) for that ad accross many sites, as well as the past performance on your page. An Ad which appears to be ON TARGET with the context (at least title and description-wise, whether the site actually has what is advertised or not), and therefore gets clicked more, may be considered "better" even though they are only bidding 1/10 what other similar ads are. So they might get 11 times the clicks (accross all publishers), but they are robbing YOU of a higher paying click. IF you are getting a CTR literally more than 10 times higher than normal (i.e. with other ads), and don't feel the visitor would click anywhere else on your site, it MAY in fact BE worthwhile. Ultimately, though if they are drawing away your readers, never to return, (I like to call them Traffic Suckers)and only paying you .03, then you are far better off filtering them and sending that same visitor to a different, likely MORE appropriate, ad which actually HAS what they advertise, and which pays .30.
How do you know which is which? Easier said than done. Usually they have taken the time to create different, highly targeted ads and landing pages for all related terms (although any good legitimate campaign would too). Check the URL for things like www.adsense-traffic-sucker.com/abc/xyz/widget-specific-landing-page... The dead giveaway of a T.S. is usually made-for-Adsense pages with little unique, new info, and LOTS of Adsense ads (or other affiliate programs - not just one) ON the landing page and no products of their own for sale. You'll also notice the better players of the game have this optimized for the highest paying keywords and bidders! They're probably making 1.00 off of your .05 which YOU could have been making in the first place! How do you determine the highest paying players in your subject area? Simply search G for your most popular keywords, G isn't stupid. MOST of the ads there are the TRUE best of the best - they may or may not all be including context matching, but it gives you a good idea.
To detect T.S.'s quickly, keep a specific channel for as many separate pages as you can and graph the avg PPC (payment per clickthru - that's earnings divided by clicks, NOT CPM which varies with CTR) day to day. If you're really determined (or mentally disturbed), or have nothing better to do (like me), record the ads running on the BUSIEST pages a couple times a day and cross reference the appearance of a new one with any dips or peaks in PPC when the channel reports are finally available. A penny or two is natural, but if it goes from say .50 to .09 in just 1 or 2 days, and only one ad has visibly changed, you've probably got a T.S.! Systematically record EVERYTHING you do, including date and time of day, and compare actions and results the next day, so that if, say filtering a URL makes things worse, you can trace back and re-instate them and eliminate another instead. Concentrate on the busiest pages first to make the quickest gains to encourage your work.
Eventually you'll know what you can expect as an optimal average CTR/CPM from each of your pages and who your good advertisers are and learn how to spot new suspicious ones instantly. Just because they're new does not mean they aren't legitimate bidders though. Always give them at least a couple days chance, they might just be your next new highest paying advertiser! Also be patient and don't do anything until GAd has settled down and settled on fairly consistent matches for your page.
It's getting to where G is either going to have to crack down on this abuse or allow us more than 200 URLs to filter, because we're getting a little close to the limit! A mininmum PPC per page option would be REALLY helpful <*hint*> :)
In the end it can all pay off with a double or better CPM which means doubling or trippling your earnings. That beats spending more time constructing new pages, getting them indexed, etc. just so someone else can make more money off of them than you are. By doing mostly just the above, along with optimizing ad placement, etc. I've quadrupled our daily Adsense earnings just since we started the program in November!
Good luck.
A mininmum PPC per page option would be REALLY helpful <*hint*> :)
Don't know if they would ever do that but there is a very nice idea I would love to see implemented!
I think in the long run it wouldn't hurt anyone except the T.S's and badly behaving affiliates (BBA's). AND in the process probably solve the TS problem at the same time. It wouldn't need to be done at all if G was basing "best ads" on the true CPM for that ad on THAT PARTICULAR page rather than for an ad overall. This is apparent because when I allow low bidding ads on a page, they do NOT generate equal or higher CPM for ME. The more I think about it, and from what I've calculated, the devious affiliates have figured out a simple way to artificially raise their OVERALL CPM and get higher placement than higher paying bids.
What if, for instance, a TS or BBA was both an advertiser AND a publisher? Which is permitted. Then they created a page similar to a very popular one they find online and decide to target with their ad. Tweeked the ad and (initially) URL filtering to get their ad to display on THEIR own page, then clicked away on JUST their own ad (using some hypothetical method to avoid fraud click detection of course - they could also do it on someone elses page or through Google search itself I suppose since I don't think there is a rule about clicking on their OWN adwords ads, but it would cost more).
Sure they would get charged for their own clicks through adwords (while making 50% of it back through their own adsense earnings) but 1. It would get their words or campaign jumpstarted and approved quickly (I'm not an adworder so I don't know the exact process or official terms) 2. More importantly they could artificially and rather cheaply raise their CPM over that of their higher bidding competitors and get better placement which would eventually get them REAL cheap clickthrus and start generating REAL earnings via their adsense ads on their landing page.
Some quick handwaving math (with MANY assumptions):
1000 self-generated adword clickthrus at .05 bid ea: 50.00
Earned back via adsense: 25.00
Total investment: $25.00
Total 'apparent' $CPM = 1000 x .05 = $50
Now 'beats' all .99 bids with 5% or less CTR: PRICELESS!
So they make a landing page showing only the OTHER bidders' ads (not their own - although there are even some advantages to keeping their own there too ;-) for .99 and above by filtering out all their own and similar low bidding ads. They'll be listed above all the .99 bids on all the OTHER innocent publishers pages. Then sit back and make .50 off of nearly every pre-qualified REAL lead that comes in for only .05 cost each. 900% leveraged profit! A mere 55 visitors and they've made back their initial $25 investment.
Actually there are a few other simple, but critical considerations I'm aware of to make it work perfectly, but left them out so as to not make it TOO easy for someone to duplicate.
Allowing google_min_bid= would help make it less profitable for these creatures, by cutting off their volume and raising the start bar and most importantly reducing the leverage effect which they depend upon. But the best way is by calculating CPM for ranking on EACH page independently and would eliminate most of the need for google_min_bid.
Hopefully now that I've exposed it G will fix it!
(either that or I've just shown a whole new generation of spammers how to make it worse for us :-).
