Welcome to WebmasterWorld Guest from 184.108.40.206
Furthermore I have disabled the displaying of adsense on these PC's and other computers connected to the internet with the same IP
Can you tell me how you did this? As I have kids in the house and the like to visit a couple of my websites. I would like to prevent the ads from tempting them to click.
I can certainly imagine a scenario where Adsense sites that have some kind of issue going on - suspect clicks, unusually high click rates, etc. - get a quick peek by a human.
Google has to strike a balance between search quality and revenue maximization. (I haven't seen your sites, Suzy, so this is a general comment and not one directed at you.) Google's reputation as the best search engine is really what drives their enterprise. If the top ten results for a keyword are scraper sites running Adsense, they might well maximize revenue in the short term - searchers will either click on an Adwords ad, or one of the organic results where they will likely click on an Adsense ad. Great for revenue, but even unsophisticated surfers will get frustrated eventually. With the proliferation of low-quality "made for Adsense" sites, I'd expect Google to start doing some weeding. A first step could be coming down hard on even minimal suspect click activity. (Or even using "invalid clicks" as an excuse - what Adsense publisher doesn't have some funny clicks once in a while?)
My solution: a modified (simpler) form of the Adsense-to-email script recently posted here.
From the internet cafe I connect to my server, my server check stats and gives me the result.
So my stats will be ONLY read from my office IP and from my server's IP. The server's IP is not used for navigation (no real users connected from that IP) so it's unlikely someone clicks ads from than IP.
Below is the script.
copy the text, save it and call it "adsense.php" or similar.
Copy it IN A PASSWORD PROTECTD DIRECTORY on your server
Should work immediatel, If your server has curl
#Insert your data here:
#Nothing to change below this
$ch = curl_init();
curl_setopt ($ch, CURLOPT_URL,"https://www.google.com/adsense/login.do");
curl_setopt ($ch, CURLOPT_SSL_VERIFYPEER, FALSE);
curl_setopt ($ch, CURLOPT_USERAGENT, "Mozilla/4.0 (compatible; MSIE 5.01; Windows NT 5.0)");
curl_setopt ($ch, CURLOPT_TIMEOUT, 20);
curl_setopt ($ch, CURLOPT_FOLLOWLOCATION,1);
curl_setopt ($ch, CURLOPT_RETURNTRANSFER, 1);
curl_setopt ($ch, CURLOPT_COOKIEJAR, $cookie);
curl_setopt ($ch, CURLOPT_COOKIEFILE, $cookie);
curl_setopt ($ch, CURLOPT_POSTFIELDS, $postdata);
curl_setopt ($ch, CURLOPT_POST, 1);
$result = curl_exec ($ch);
[edited by: frox at 4:11 pm (utc) on Mar. 24, 2005]
trillianjedi: I am not in the habit of checking my stats from others computers. I have done it a few times (from my moms house).
dirkji: I showed adsense on several sites. One of these was my "breadwinner" so to speak and the rest accounted for a small percentage of my adsense income.
webwork: I actually started webmastering in the adult realm, and have moved entirely into the mainstream realm now. However, i still sympathize with the adult webmaster, so to speak, and i see how a program that allowed adult sites could be a win win situation for adult publishers and yahoo (or whoever).
Also, like I said, I have been doing this since 1999 and i have done what you may call typosquatting before. It was before I had ever heard the term "typosquatting" and I just thought it was a really smart idea. At the time, as far as I know, it was something that had rendered itself as a shady practice and it seemed to me that i was helping the companies whose traffic i was collecting. This was never a major part of my business, and I no longer have any "typosquatting" urls.
You do raise a point though, that I have had really weird adsense days. Oddly, these were more when I first started using adsense and my traffic was erratic (since i was adding/testing adsense in a variety of places/site). At that time i would experience fluctuations of $x to $5x dollars on the same number of daily clicks. For the last 6 weeks or so, my income, CPM etc has actually been really steady--with a slow decline of CPM. So, it doesn't seem like this is the issue--but, it is a good point, maybe i should have contacted them to demonstrate my good faith when i did have weird days.
I am going to write another post about my sites since this one is getting long.
...why can't Google just limit the number of ad clicks it will "accept" from a computer/IP address to a certain number per day (say, 2)? There's got to be a better system, since a competitor can so easily destroy another site's account.
some of my sites provide materials that are usefull in schools. there could easily be 100 students hitting my site with the same ip in a given day.
To address the question of my sites. The main "breadwinner" site I have is undoubtedly a useful content-filled resource. It is a smallish site, because it is a pretty small niche, but it has info and articles about everything you could want to know about that little niche.
I do have some other domains that I have put adsense up on that are less clearly content-filled. A few of my sites are type-in domains that are not clearly connected to a marketable product. In this circumstances, I put up a one-pager with an article (well-researched) about the subject, I also put some handpicked links to other resources with more info about the subject, and I put adsense ads. I design these sites thinking about if i personally were looking for info on those widgets, would this be a useful place for me to land? Though these are one-pagers, I make them useful (in my opinion). However, somebody else could see it differently. It is worth noting that these one-page sites make only about %5-%10 of my adsense income.
My adsense email said my account was terminated due to "invalid clicks". Do the termination emails always say "invalid clicks"?
[edited by: freeflight2 at 5:04 pm (utc) on Mar. 24, 2005]
Do the termination emails always say "invalid clicks"?
Probably since if someone broke any of the programs TOS - any clicks that were received on the site would be invalid.
You stated that you built single page sites and put AdSense on them.
Where these big $ terms/subjects as far as AdSense/Adwords?
Would you have built the pages and researched them if it wasn't for AdSense potential?
These single pagers were not big money terms by any stretch, and I NEVER built them with the idea of targetting any term etc.
They were type-in domains, most of which i have had for years. I have tried different stuff with them over the years: redirecting to affiliate programs, domain parking pages etc...but, since they aren't really particular consumer product related, they never made much money. They still didn't really make much money...
As far as the research, i had already done much of it before adsense existed in my efforts to try and best understand how to monetize these terms.
Honestly, I may not have put the particular format of an article and handpicked links on the pages if it weren't for adsense. But, I did not aquire or build the sites with adsense in mind.
I see how it can be a slippery slope...but, i am/was pretty confident that i was not violating TOS and that i was providing valuable content...
These single pagers were not big money terms by any stretch
The money makes no difference - you are either in breach of the AS TOS or not.
I'm not sure that you were actually in breach of the TOS having ads on these doorway and typo pages, but it certainly wouldn't reflect well on your quality as a publisher (unless those pages converted like crazy of course) on a manual/human review.
lammert: I may do this at some point (put my url up), right now i am suffering from uber-paranoia about things, and want as few people as possible to know who i am relative to what sites etc. I am totally afraid that it is something personal, either malevolent (competitor) or benevolent (dumb friend who thought they were doing me a favor).
asp4bunnies--thank you so much! I really appreciate you taking the time to outline your experience and offer your helpful advice.
I can inagine that an effective test for invalid clicks would be a comparison/correlation between network-wide CTR for a set of ads and a site's CTR for that same set. Especially when the set was artificially chosen to vary from the site's normal CTR pattern.
To exemplify: If a site normally shows a 1.5% CTR, G may occasionally alter the ad mix so as to show a set of ads that network-wide deliver more like 0.5% CTR. Once the site passes the test, normal targeting resumes. If the CTR remains high, then it's higher than expectations, and further investigation is warranted (leading eventually to the dreaded email). Even a really smart clickbot couldn't be aware of the network-wide CTR for an ad, or a set of ads, so the lack of a statistically significant change in CTR following a change in ad selection like this would be good (maybe even excellent) evidence of invalid clicks.
I'd expect something like this (maybe I have some of the nuances a little wrong, but a statistical approach in general) to appeal to G since it can be pretty highly automated--set a threshold for change, then if a site does a positive change over the threshold, give them the above validity test. It should give a low false positive rate, and they can fine-tune their triggers with a lot of data that we never see.
And, fully recognizing that anecdotal evidence is not really all that valuable, I think I've seen it in my own account more than once. Most recently, I relaunched two of my topics as subdomain sites, with an entirely different template, different colors, different ad format and placement. CTR improved. Imps improved as I promoted the relaunch. Earnings improved. I hit about double my 30-day moving average earnings, and then CTR fell markedly (with earnings also). I couldn't see a huge targeting difference, but I only have a narrow view of the situation, so that's not that meaningful.
A couple days of that pattern ensued, I gritted my teeth and did nothing, and then my CTR popped back to a level similar to before. I acknowledge that this is purely feckless speculation, but I think I was undergoing a validity test (seemingly I passed).
I'm also by no means suggesting this is the reason behind long-term drops in earnings like some of the 90% drops others have reported. I do think that some of the variation in stats is due to testing for invalid clicks, and I do think that sites who pass get "paid back" with a small earnings spike shortly after a test like this. I can find that pattern a few times in my stats anyway. I'm curious if any of you can see something similar.
I noted that Suzy did report a drop right before the dreaded email arrived.
You will not receive any check that is currently unprocessed
It writes like this on the post. I know my february earnigs have been processed and I also think that they must have been given to mail by now. And in the tos it writes if the account is cancelled they will send you the money if its over ten bucks if i am not mistaeken...
I can even imagine that a statistical test is done within a site by comparing page statistics. I always closely monitor my channel data. I have some channels with very constant CPM due to high CTR and ads with near to stable value, and other channels where CPM is fluctuating heavily because of some ads with high value and other ads with low value.
Every page with ads has its own characteristics and clickbots, or people artificially clicking on them do not know these characteristics .It is impossible for them to simulate clicking on ads in the same way as natural clicks occur.
you mentioned blocking all Adsense blocks from appearing on any computer on your local area network... how might I go about doing this myself?
I use serverside include:
<!--#IF EXPR="$REMOTE_ADDR!=10.0.0.1" -->
<!--#INCLUDE VIRTUAL="/includes/google.inc" -->