| This 255 message thread spans 9 pages: < < 255 ( 1 2  4 5 6 7 8 9 ) > > || |
it wasn't me!
| 8:29 am on Jul 1, 2003 (gmt 0)|
Hi all, I got this email from Google Adsense saying that they'd detected fraudulant clicks with my account and that they were investigating!, but in complete honesty it wasn't me!
I've only had the ads on my site for a few days now + may have made 2 or 3 test clicks on the first day, but that was it.
Do you think they're trying to get out of paying me?
| 6:19 pm on Jul 21, 2003 (gmt 0)|
>>* A webmaster wants to check the advertised links, to see what their content is, so he will be able to ban the links, so his visitors will not get links to 'evil' sites which for example sell stuff way too much overpriced. <<
1. Mouse over the link and copy the link.
2. paste it into the address bar, just leaving the actual url and removing your id and other info
3. click go!
It may take a bit longer but you owe it to the advertiser and if you want to check links you should be willing to go to the extra effort to make sure the advertiser isnt paying for you to "review" their link.
| 6:36 pm on Jul 21, 2003 (gmt 0)|
Yes something like that is not a problem for me.
But I'm more worried about fraudulent clicks made by other people. I mean there are always people or webmasters with a site about the same subject as your own site, who will not like your site, and if they know how they can **** it up for you, some will do.
And how will Google check this?
| 6:42 pm on Jul 21, 2003 (gmt 0)|
>>And how will Google check this? <<
Just a guess. They could email and ask you if the domain where the clicks are coming from is owned by you. im preety sure they record that. Or they could check whois and ownership info on that domain. Not perfect, but will go a long way to proving other's cunning stunts.
Remember google has heaps on info about you and others. cookies, toolbar data, whois, a history of your sites, and a full profile of your site that goes far beyong what you can check yourself. I guess they have ways to solve this problem to a large extent.
| 8:35 pm on Jul 21, 2003 (gmt 0)|
I'm not sure whether Google will take the time to discuss matters. They are more likely to take action and await your query to them.
| 4:08 am on Jul 22, 2003 (gmt 0)|
I got the same dread message late last month, right before I went on vacation. I emailed them very politely and said Please Sir, I wasn't defrauding you, Sir, I maybe tested two or three ads, Sir. I'm an honest hardworking webmaster with really targeted traffic.
I also asked my Adwords account rep to put in a word for me.
While I was away they reinstated me, saying that a fraud filter had incorrectly flagged some clicks as fraudulent. (Fortunately I had left the code in place so things picked up where they had left off.) No further details. I have been seeing double digit click through rates, though, which could have triggered something.
My previous balance was also reinstated.
Now I just need to figure out how to add some zeros to my daily earnings and figure out which beach to check in from...
- plead your case and ask nicely for them to review your case
- if you are an adwords advertiser, tell them so
- if you're not being evil, tell them so
- if you have any contacts at google, do the same with them
| 4:12 am on Jul 22, 2003 (gmt 0)|
How can they detect fraudulent clicks? Not sure how. But here's a real simple way they can detect innocent yet misadvised webmaster clicks: match the IP/cookie of adsense admin log-ins to the IP/cookie of adsense clicks.
In fact, that one is such a no-brainer I don't know why they don't have a filter on there for that to automatically disregard owner-operator clicks.
| 4:16 am on Jul 22, 2003 (gmt 0)|
Probably, some publishers are mobile, even using different computers/access.
| 4:19 am on Jul 22, 2003 (gmt 0)|
welcome to the world of publisher fraud. a factor that has haunted the non-search ppc community for 6 years now. and only one variable amongst many that should signify to us that it is unacceptable to pay the same cpc for a serp ad and a contextual ad. there was no doubt that this was going to be a problem, and we should not be surprised when flags for redundant clicks go off. ADsense has a can of worms in its hands because it is a catch 22~ the more you grow your non-search ad serving base, the more vulnerable you become to publisher fraud ~ it permiates all ppc networks. and, one should not underestimate the degradation of quality of ADsense relative to serps ~ charging the same cost per click is a play on our ignorance that we should not tolerate and publisher fraud is just one of the reasons why.
| 6:18 am on Jul 22, 2003 (gmt 0)|
|charging the same cost per click is a play on our ignorance that we should not tolerate |
AdWord/Adsense rates are set by bidders, not by Google, and the market will determine the value of a click.
| 7:04 am on Jul 22, 2003 (gmt 0)|
|How can they detect fraudulent clicks? Not sure how. |
At first I thought fraud would make AdSense unworkable, but after more consideration I now think fraud on any serious scale will be damn difficult.
Remember that Google has a huge amount of data - it has records not only of every clickthrough, but of every ad impression as well, and all the IP addresses and page URLs to match... To fake clickthroughs without creating any kind of pattern that will show up, a thief would have to create fake impressions as well, distributed over enough computers to match the overall pattern of the AdSense data. Imperfect matching of that would limit how much they can steal to some proportion of how much they would have got legally (so they'd be risking an above-board $100/month to steal an extra $20, say); and even with perfect pattern matching they'd be constrained by the number of computers they can crack into.
But there may be criminals out there with PhDs in data-mining, control of tens of thousands of computers randomly distributed around the world, and a number of quality high-traffic web sites they're prepared to risk.
Anyway, if you get approached by someone offering to "double your clickthroughs" for payment, don't post here, go straight to Google security.
| 7:29 am on Jul 22, 2003 (gmt 0)|
If you are talking of large scale crime, syndicates will recruit, procure and operate in such a way thay it would cost billions before the system was stopped. Before that time there would be bases all over the world, recruiting, procring and operating.
Take credit card fraud, it is an industry by its own merits, On the 'clean' side there are companies making millions from holographic printing methods, chips, plastic cards, EPOS, data capture .... etc etc, this is just for fraud prevention. On the 'dirty' side - see the above.
I think if it does reach the same scale, Google will be a busy girl!
| 7:29 am on Jul 22, 2003 (gmt 0)|
chicago is right on PPC fraud in general. its alway a can of worms. However adsense has some advantages. It not only had a lot of data on publishers and advertisers which other ad networks and PPC providers have, but also a mass of historical data on domains, PLUS toolbar data etc. Yes, now OV/ATW/AV/Y! has this most of this now too, but not on such a broad scale.
| 9:32 am on Jul 22, 2003 (gmt 0)|
I've had the same problem today. Came in to work to find an email from Google saying "It has come to our attention that fraudulent impressions or clicks have been generated on the ads on your site..."
I've never tested any of the ads by clicking on them A) Assuming that Google know what they are doing and B) Not wanting to take the risk of getting thrown out of a well paying program.
However, I do have a high clickthrough rate. It's over 20%. I put Adsense on a few days ago and it hovers between 20 and 23%. Those are honest click-throughs guys. Promise. And unlikely to be one mad visitor clicking frantically as the % of clickthroughs is consistent 24x7.
Judging from some of the comments here some webmasters are getting well below 15%. Now, I'm not complaining about the (genuinely) high click-throughs - and the earnings - but how do I convince Google that I'm an honest guy and those are genuine? Also, if there is someone upset with us enough to keep clicking on the ads to annoy Google, how do I convince Google to do the correct research to identify them and/or discount those clicks without throwing us off the program?
Sorry, I'm deleting some text and figures as I'm told this is against the forum policies
[edited by: Poweroid at 12:14 pm (utc) on July 22, 2003]
| 9:52 am on Jul 22, 2003 (gmt 0)|
Don't Google open any form of dalogue after these 'it has come to our notice emails?'
| 10:32 am on Jul 22, 2003 (gmt 0)|
Good to hear your issue has been completely resolved!
I, too, responded with a polite email explaining that I had done nothing wrong, and asked them if certain factors could pose a problem (e.g., IP of my server and my development machine being the same).
I got an answer from them that addressed my questions --so, YES, they do open a form of dialogue!--, and was told that my account would remain active.
I would *love* to get that "fraud filter had incorrectly flagged some clicks" message, though, so that I know that the issue has been solved completely; but for now I don't really worry --I *know* that I have done nothing wrong, and AdSense is doing quite nicely for me...
>> match the IP/cookie
Unfortunately, that does not always work for two reasons:
1) People behind a proxy may *share* the same IP address (I understand this applies to AOL users in particular)
2) People on a dial-up connection typically get a *different* IP address every time they dial in (dynamically assigned by the ISP from a pool of then-available addresses)
>> charging the same cost per click is a play on our ignorance
If by that you mean that CPC should *not* (necessarily) be the same for GG SERPs and AdSense network partners, you may have a point --I'm not an AdWords advertiser myself, but if the conversion/"value" of clicks from one traffic source is much lower than the other, I would not expect to pay the same for both sources... If there is no way to set different CPC, your only option may be: not showing up on any AdSense network pages *at all*.
If your clicks are legitimate, then I would not worry about it at all. I'm sure GG *will* do the required research, and will treat us fairly!
| 10:52 am on Jul 22, 2003 (gmt 0)|
Thanks Eltiti, I hope so :-)
It's also heartening to know that they do talk to you and that a person actually manually looks at what's been flagged.
| 2:22 pm on Jul 22, 2003 (gmt 0)|
|Chicago..If by that you mean that CPC should *not* (necessarily) be the same for GG SERPs and AdSense network partners, you may have a point --I'm not an AdWords advertiser myself, but if the conversion/"value" of clicks from one traffic source is much lower than the other, I would not expect to pay the same for both sources... If there is no way to set different CPC, your only option may be: not showing up on any AdSense network pages *at all*. |
exactly ~ Eltiti.
| 3:16 pm on Jul 22, 2003 (gmt 0)|
It looks more and more people got the red flags from Google INCORRECTLY and/or CORRECTLY but this does concern me a little about how effective Google's fraudulant detecting system.
By now, I only saw publishers "cleared" their "fraudulant" claims. Why does Google so "simply" send out the red flag notice? They should work "secretly" at the back-end and "setence" those real "dirty" guys without doubt.
Google should broadcast "By now, we have removed XX publishers from our Adsense program due to the fraudulant actions". This should help advertisers to establish confidence and have effective guide to other publishers.
I would definitely keep one eye on this issue!
| 4:28 pm on Jul 22, 2003 (gmt 0)|
>>Chicago..If by that you mean that CPC should *not* (necessarily) be the same for GG SERPs and AdSense network partners, you may have a point --I'm not an AdWords advertiser myself, but if the conversion/"value" of clicks from one traffic source is much lower than the other, I would not expect to pay the same for both sources... If there is no way to set different CPC, your only option may be: not showing up on any AdSense network pages *at all*.<<
Our own experience as an adwords advertiser is that we get less CTR from "content site" ads, they are slightly more expensive, but there is good evidence they are CONVERTING even better than google SE clicks. I know this seems strange but from asking our leads where they come from, this seems to be the case.
In short then, Im not at all convinced that content ads are less useful re conversion into leads and motivated buyers, in fact the opposite. On the other hand so far they are slightly more expensive (remember it is not the SAME price, it depends on a variety of factors when you have non exact keywords in Google - without the apostophies or brackets), and the CTR is less. But of couse we dont pay for impressions, just clicks, and they are not counted in CTR for keyword exclusion purposes.
This has only occured since Adsense started, before the hots from content sites were not as useful, i guess due to them being on some broad topic based mega-sites only..
Therefore if there was any move to having diff costs for content vs search sites, it would make sense from where i stand to pay MORE for content ads. At the moment i think its better not to play too much with this...
I'm sure people's experience may differ, depending on your industry and the type of sites you are appearing on. In our case we sell business consulting and research services in a very niche area.
Gets more interesting everyday...
| 4:59 pm on Jul 22, 2003 (gmt 0)|
>>pay MORE for content ads
i don't like disagreeing with *you* ~ but Chiyo, that is a very very very tough pill to swallow. reactive ads can not equal proactive ads from a user behavioral standpoint. it is just not possible.
the topic of fraud is just one (non-behavioral) of many many (behavioral) reasons why. most important is the intent of the prospect.
i dont want to rehatch these arguments again, however. i try to stay out of adsense threads and instead just read how it is working for you. but with comments like that, Chiyo....
| 6:21 pm on Jul 22, 2003 (gmt 0)|
|i don't like disagreeing with *you* ~ but Chiyo, that is a very very very tough pill to swallow. reactive ads can not equal proactive ads from a user behavioral standpoint. it is just not possible. |
I assume that your definitions are:
Proactive ads - Ads that a user sees while searching on a topic
Reactive ads - Ads that a user sees while browsing
But there's also a third category: Proactive ads that a user sees after finding information in a search. It makes perfect sense that these would have a higher conversion rate than an AdWord on a Google SERP, because the reader has:
a) Searched for information on the topic in Google.
b) Clicked on the ad after reading the information that he found.
Let's take a hypothetical example of a person who's interested in a memory card for his Widgetco D-1 digital camera. He knows he needs a bigger Compact Flash card, but he isn't sure what type to buy, so he does a Google search.
Now, if he sees an AdWord for Compact Flash memory cards on the Google SERP, he may click on it just to compare specs. But if he clicks that same AdWord in the AdSense "Ads by Google" box on a photographic site's page titled "How to buy memory for your digital camera," he's much more likely to be a buyer. Why? Because the article has answered his questions about what kind of memory he should buy for his camera (a 1 Gb Type I Compact Flash Card with write-acceration memory). By the time he clicks the ad, he isn't a person in search of information--he's a serious buyer.
IMHO, it's naive to lump "content sites" together, because different content or information sites will have different characteristics according to the nature of their audiences, whether they're editorial or community sites, and so on. Getting back to our digital-camera example, a page about digital cameras at HowStuffWorks.com will have a much lower percentage of "serious clickers" (and a far lower conversion rate) than a page about digital cameras at DPreview.com or Imaging-resource.com. And an article on Caribbean cruises in THE NEW YORK TIMES will attract more casual tire-kickers than an article on Caribbean cruises at CruiseCritic.com or CruiseDiva.com.
Smart advertisers don't make broad assumptions about what works and what doesn't work. Instead, they test. They also use good judgment when writing ad copy. For example, an ad that says "20% off Silversea Cruises" may work fine on a cruising site or in a review of a Silversea cruise, because the reader will know that Silversea is a luxury line with per diems in the $700-800 per day range. But if the ad is going to appear in a wide variety of media (including Google SERPs), it might be wise to include a range of fares in the body copy (or at least a minimum fare) to discourage clicks by first-time cruisers with limited incomes who may not know that Silversea costs a lot more than mainstream lines like Carnival and NCL.
| 6:53 pm on Jul 22, 2003 (gmt 0)|
Well, I've read the entire thread - and some of it has me a bit concerned. I see that a few advertisers have posted in here - stating that they don't want webmasters clicking the ad links on their own sites, and I can surely understand that they would not want to pay for advertising when the person clicking does not intend to purchase anything ...
However, let's say I go to Google.com and search for something I want more information on (let's say "web hosting") -- a bunch of ads show up to the right, and the short description catches my eye - I click on it - look around in the site - and don't find what I'm looking for ... Now, considering that my site deals with web design - I get very similar Adsense ads on my site. I spend a lot of time at my site, and it has a high traffic rate. Now, what is wrong with me clicking on one of those ads - if it has something in it that catches my eye. Why should I go through a big workaround just to avoid clicking on an ad (like I said above, the same thing would happen if I simply searched for something on Google).
I guess my point is ... I'm a consumer - just like anyone else ... Placing those ads on search results at Google - or on my website - doesn't guarantee that I (or anyone else) will actually buy anything at the advertiser's website. But, I pay for services - I buy products - and I do most of it on the internet. I've clicked on a few of the ads on my site - and I don't see anything wrong with it. I've come across sites that I didn't know existed - and I've bookmarked a few of them (places where I can buy software at discount prices - and that is something I find very useful).
I'm certainly not promoting click-spam or anything of the sort - that is not my intention. I'm just a little concerned by the outrage that a webmaster might click an ad or two (or three or four, if the ads are of interest).
I could do a search for something on Google, and it might produce tens of thousands of search results -- and I may never find what I'm looking for. However, sometimes those ads can help ... I think these ads serve their purpose - for both the advertisers and for the consumers -- and once again, I am a consumer (not just a webmaster).
again, not promoting click-spam - just trying to make some sense (which I probably didn't accomplish) :)
| 7:35 pm on Jul 22, 2003 (gmt 0)|
my old mate chicago wrote >>i don't like disagreeing with *you* ~ but Chiyo, that is a very very very tough pill to swallow. reactive ads can not equal proactive ads from a user behavioral standpoint. it is just not possible.<<
Ahha! but what you are talking about is theory and im talking research and practice! Its what im seeing..
Theories are good stuff, they organize and make order out of lots of past observations. But sometimes they need revising when the situation changes or is different.
As i said, what I'm seeing may be different from what others are seeing. I'm just questioning the theory since i found a black crow. ;)
| 7:49 pm on Jul 22, 2003 (gmt 0)|
|Smart advertisers don't make broad assumptions about what works and what doesn't work. Instead, they test. |
I wholeheartedly agree with that statement; I think it applies to smart publishers as well.
My own "intuition" was similar to Chicago's --when I said: "one doing better than the other", I actually assumed that SERPs would do better than content ads.
But when I read chiyo's post, I was reminded that it is usually best not to "assume" too much, and test instead.
The scenario you describe certainly sounds plausible; but so does Chicago's general assumption. What you describe may only be true for certain products, or it may apply to all products in a certain category; or...
OK, so we need to test. But in order to test, we need data.
As I mentioned before, I'm not an AdWords advertiser, just an AdSense publisher. I do not know, therefore, exactly what data/statistics you get with AdWords; I do know that the AdSense stats are of very limited use to a publisher trying to optimize.
I have ads on two of my sites, one about widgets, one about gadgets. I don't know how many clicks are generated by either site (I don't even know that for the impressions, unless I look at my own server logs!); and I would need to know it at the page level, not the site level, in order to "experiment" and optimize my revenues...
Oh well --this has been brought up in another thread before, and Google has indicated that additional stats may be available in the future. But for now, I think it is safe to say that, in economic terms, the aggregated data currently available will lead to a sub-optimal outcome.
| 7:53 pm on Jul 22, 2003 (gmt 0)|
|what you are talking about is theory and im talking research and practice |
My own post was so long because I didn't have time to make it shorter... ;) What you and efv state, captures the essence nicely!
| 8:40 pm on Jul 22, 2003 (gmt 0)|
Chiyo and EuropeforVisitors, eloquent as always.
Your take: Theory vs. Practice.
My take: The Norm vs. The Exception
EuropeforVisitors, your scenerio of a "third type" is quite compelling but unfortunately too perfect, and most certainly not the rule. My CTRs for ADsense consistantly run under 1/2 of 1% (even with publisher (and friends of publisher) clicks). PPCs on SERPs generally range from 6-15%. This is another significant argument because, CTR portends the behavioral disposition (intent) of the user relative to the ads that are being seen.
The argument that it doesnt matter, you only pay for clicks, and the clicks you get are more qualified, once you get them, is flat because it doesn't accurately differentiate between the motive of a contextual clicker relative to a clicker on a serp. in other words, in EFV's perfect scenerio, one finding a content site through a serp, clicking and reading, and then finding a contextual ad, does not account for others who clicked who did *not* find you (the publisher) that way ~ quite easily the majority ~especially with a content driven-sticky site.
Chiyo and Europeforvisitors, the two of you are exceptions to the rule and without question smart enough to take a general theory and prove it wrong. There is no doubt what you are seeing Chiyo is real. Unfortunately most of adsense will not be you. Especially as this gets bigger and bigger and bigger...
I think ADsense is great for advertisers and publishers ~ i simply would like to see separate systems. Thereby accounting for the unique charecterisitcs of each, and allowing for an isolated ROI and an accurate picture to judge PPC bid value, seperate from that of PPCs on SERPS.
Regardless thanks for your valued input on your use of ADsense, it is helpful to us all.
| 10:39 pm on Jul 22, 2003 (gmt 0)|
|The argument that it doesnt matter, you only pay for clicks, and the clicks you get are more qualified, once you get them, is flat because it doesn't accurately differentiate between the motive of a contextual clicker relative to a clicker on a serp. in other words, in EFV's perfect scenerio, one finding a content site through a serp, clicking and reading, and then finding a contextual ad, does not account for others who clicked who did *not* find you (the publisher) that way ~ quite easily the majority ~especially with a content driven-sticky site. |
Well, you don't necessarily know the motive of a clicker on a SERP, either. That person may simply be looking for product specs and information (especially if the top SERP pages are cluttered with sites in Japanese and several languages other than English, which was the case when I searched on a certain digital camera a few minutes ago).
As for a "content-driven sticky site," a content site isn't necessarily sticky. A destination-oriented travel site, for example, will be less sticky than a site like CNN.com or The New York Times because most people are likely to visit only when they're planning a trip. And a site that *is* sticky, like a photography hobbyists' site or a ham-radio site, may cater to people who have a greater likelihood of buying than the average searcher does. It all depends on the site and the audience.
I agree that advertisers' concerns need to be met, but I don't think that dividing the AdWords/AdSense world into "search ads" and "content ads" is the answer. What's really needed is a greater range of options for the advertiser: for example, the ability to include or exclude by domain, by type of page (editorial, blog, forum), or perhaps even by category (based on keywords that are associated with the domain in the publisher's account information). This would make it easier for advertisers to track and test results, and it would lead to greater advertiser confidence in "content ads."
BTW, one thing that hasn't been mentioned in these discussions (but which needs to be considered) is what the advertiser is trying to accomplish with the ad. Not every advertiser is looking for a profit on every clickthrough. Some advertisers--especially large brick-and-mortar companies, as opposed to small merchants or affiliate sites--may be more interested in acquiring customers or obtaining leads than they are in an immediate ROI. For such advertisers, a clickthrough from a "sticky" special-interest content site may be worth more than a clickthrough from a SERP because of the user's demonstrated interest in the topic. It all depends on the site, the topic, the audience, and the advertiser's goals.
| 11:31 pm on Jul 22, 2003 (gmt 0)|
|Well, you don't necessarily know the motive of a clicker on a SERP, either. |
they might even be "the competition" clicking your ads to run up the bill.
having ads on a publisher's site probably hides it better from competitor's malicious clicks.
do it once and you'll get away with it, do it a 100 times and google will catch it.
| 12:38 am on Jul 23, 2003 (gmt 0)|
hooloovoo22 - I know my competitors come to my site daily and am sure they will click on some of the ads as they will be interested to see who is advertising and some of them will be even more surprised to see it is themselves.
The serious competitors may well click once or twice for reasons mentioned above but as EFV has mentioned in other threads they will not click the links 'to try and get the site in trouble' they have much better things to do with their time.
I would like to comment on what Chiyo, Chicago and EFV have been discussing but will finish my first coffee before I do!
| 3:35 am on Jul 23, 2003 (gmt 0)|
Did Google actually suggest that YOU were the person committing fraud, or that your account might have fraudulent clicks? To me, this is an important distinction. I bet a person could go to some publishers web site and click on a link a whole bunch of times. Is that the publisher's fault?
Coincidently, I got an email today too. I guess they are trying to "crack down" on their publishers. The thing is, the page they listed in the email did not even have a google adsense banner on it. I think a lot of it is robotic.
Anyway, I can't wait until Google has competition in this arena. I HATE being at the mercy of one company. Perhaps Microsoft will buy them and I will not have to worry about making money.
| 4:54 am on Jul 23, 2003 (gmt 0)|
>>Did Google actually suggest that YOU were the person committing fraud, or that your account might have fraudulent clicks? To me, this is an important distinction. I bet a person could go to some publishers web site and click on a link a whole bunch of times. Is that the publisher's fault?
I put Google Adsense on 5 of my websites, for about 20 days. All was well for a while, not a real impressive CTR, but it did make $1,500.
Then I got an email saying, "Thank you for your interest in Google Adsense, Thank you for your interest in Google AdSense. Our program specialists
have reviewed your application. Based on our Terms and Conditions we are unable to accept you..."
I replied to Google, nothing. I wrote again, nothing. Finally I got an email saying that they had detected fraudulent click activity and emailed me asking me to take measures to prevent fraudulent click activity. I never got the email and seriously doubt they ever sent it; if if they had, what can I do to stop people from clicking on my ads?
Furthermore, I seriously doubt they did detect fraudulent click activity. Nobody has the time or patience to click on ads like that.
Anyway, Google did succeed in ripping me off to the tune of $1,500, but that's life. You expect it from most affiliate networks, and Google is no different.
I seriously think Google needs to mature in this area. Overture has known for ages that people will click on their competitor's ads so they just don't count those clicks. Google would do good to follow suit.
| This 255 message thread spans 9 pages: < < 255 ( 1 2  4 5 6 7 8 9 ) > > |