Click Fraud Suit Filed Against Search Engines [Archive]

Mikkel deMib Svendsen

04-06-2005, 05:48 AM

I know it is called "click fraud", but where is the fraud? is it the SE for not discovering the "fraudulent" click? Is it the clikck generator? Who?

There are many kinds of fraudulent clicks. Some of them comes from identifyable bots and IPs know to be almost pure fraud. We all know that - and so do the engines. The question is, if they filter out ll clicks they KNOW comes from such fradulent sources. If not, then I think they are guilty.

does any independent group monitor this??

Yes, I believe news paper distribution is monitored - at least here they are :)

There are many other areas where the merchant have the full responsibility - for example credit cards. If I go and tell my bank: I did not buy that item, THEY have to prove to me (with the help of the merchant) that I actually did. If they can't prove it (and even sometimes when they can) I keep my money.

An in fact, this is one way that lawyers has been talking about taking the click fraud cases - through the credit card laws :)

dannysullivan

04-06-2005, 07:34 AM

We're working on getting a copy and will post to the blog when we do. In general, people do ask for refunds. The complaints often are that they don't get them or for as much as allowed. The fraud part is that someone is costing someone else money for their own gain. Competitors are having clicks driven up to cost them money. Others are clicking on contextual ads just to make publishers money. I think the real issue ultimately won't be is there fraud. There will always be some type of fraud. But have the search engines been found to full work to do what they can to prevent it and refund when found. There have been stories that they've allowed fraud situations to continue, even when they've known. If that actually was proved in court, not good.

I, Brian

04-06-2005, 08:23 AM

M$N is noticeably absent from the list.

They don't have their own advertising network. Not yet.

As for clickfraud - this is an issue that's been bubbling up for a while. A couple of short reads:

Mike Grehan's The click, the fraud and the ugly side of search (http://www.netimperative.com/2005/02/01/click_fraud)
John Battelle's Other Shoe on Keyword Prices, Clickfraud (http://battellemedia.com/archives/001296.php)

Also, as covered on SEW:
Are we all getting ripped off? (http://forums.searchenginewatch.com/showthread.php?t=143)
Click-Fraud said to be 50% of clicks (http://forums.searchenginewatch.com/showthread.php?t=1887)

but especially:
Anti Click Fraud Tools (http://forums.searchenginewatch.com/showthread.php?t=3384)

projectphp

04-06-2005, 09:25 AM

The article is scant evidence, but the article says:
...allege that the Internet companies knowingly overcharged for advertisements they sold and conspired with each other to continue doing so.
Now, the "Knowingly overcharged" bit is interesting, and I assume it means on a total bill, not by keyword (i.e. not that they charged $077 per click rather than $0.55). The collusion is even more interesting. Is that supposed to mean that Yahoo, Google et al got together and agreed to keep doing this? Or that they conspired with publishers to continue to do this (partner sites, e.g. Ask and Google)?

The fraud part is that someone is costing someone else money for their own gain.
I looked up fraud, and that is pretty close to the definition I got (well done :)). But I guess I just wonder if it is legally fraud on the engines part? Are they responsible to the extent that it is fraud that they are perpetrating?

To prove so, one would need to show that PPCs knew specific clicks were fraudulent, or should have known, and continued to bill advertisers for these clicks. So, it all rests then on at what point should PPCs have known?

Some other (quasi-random) thoughts:
1. What defines a "click"? To prove there are fraudulent clicks, you would first have to define what a click is and, conversely, isn't. The model is "price per click", or "Cost Per Click", but does the word "click" imply that it should involve the pressing of a mouse button, or is click defined in terms of an HTTP request from a computer to the PPC engine? The engines could theoretically define click as whatever they like. https://adwords.google.com/support/bin/answer.py?answer=6382&ctx=en-uk:top5 is one page on which Google uses the phrase "click". does anyone know anywhere where Google or Overture define "click"?

Really can't see the engines arguing this, because it would really undermine trust in the service, which is worth more than refunding all the money in the suit most likely, but it is still a consideration. A good thing out of all this might be, if the case goes to court, that we may end up with a definition of what "click" really means. This is especially true for the click fraud "fringe", for example the multiple clicks in one time frame of X. What is acceptable, i.e. could be a human clicking multiple times, and what is fraud? Is 100 clicks a day fraud? Is it 10, 1 a day every day, 3 a week? How many? At what point should a search engine be legally required to act and issue refunds and what defines a legit or fraudulent click will be interesting.

2. What auditing does / did Google et al do on clicks, and has Google, Overture et al ever given a refund/s without being prompted by the advertiser? As Danny said, we all agree that some fraud does happen (even though I feel there is some dispute as to what defines a click). Given some fraud is a given, the engines must, with any reasonable amount of due diligence, have known about at least some of it. What action the collective engines did (or did not) take refunding this ill-gotten booty is absolutely vital. If I was Google, Overture or LookSmart, I would want to have refunded, "unprovoked", at least some money, even a few thousand out of the millions collected, well before the suit was filed, otherwise I would be feeling ever so slightly scared by the suit.

3. Along the same lines, there are known cases in which AdSense publishers were "sacked" for "bad" clicks (you couldn't see for the other threads @ WMW some months with all the AdSense complaints). The removal of some sites from the publisher network is, IMHO, both a plus and a minus for Google (alone on this occasion). I thought, at the time, that it was dangerous to sack publishers without issuing refunds / asking for funds to be returned. After all, if there were reasons to sack people, there must have been fraudulent clicks, ergo refunds to advertisers should have been on the cards.

DISC

04-06-2005, 09:53 AM

The various financial motivations behind click fraud – for affiliates and publishers, competitors, disgruntled former employees, or even SEM consultants who have not been paid by a scurrilous client – are fuel enough for click fraud and subsequent law suits. However, I feel that not enough attention has been paid to the motivations behind virus propagators, motivations that pose a huge threat to the PPC industry. What financial incentive is there for a 14 year old hacker to program and release a virus? Though I’m not a programmer, I can imagine that it would not be too difficult for a hacker to propagate a script that unleashes torrents of bots that search and click PPC ads, perhaps with a set of phrases that attack a given industry. Such bots could use rotating proxy servers, randomized timing, and other methods that would render millions of dollars worth of click fraud over weeks and months virtually undetectable as fraud. Not only a hacker, but also a zealous terrorist could do willful damage to a major Western capitalist industry in this way. I have no idea how or whether such an event would play out in the legal system, but I worry that within 12 months or so, a major attack like this will happen.

PhilC

04-06-2005, 10:10 AM

That's not hacking - it's programming, and it's already happened. Programs do make the clicks, and have done for years.

DISC

04-06-2005, 10:24 AM

That's not hacking - it's programming, and it's already happened. Programs do make the clicks, and have done for years.

True enough. But the larger concern still stays with me, regardless of how we lable who does it.

krisval

04-06-2005, 11:59 AM

Is any of this actually illegal? I mean, is it against the Google T&C to do so? Do they promise to deliver only "human" clicks? And only so many from any one human in a day?

ABSOLUTELY I had an issue with a second tier company last year and they refused to refund my money until I actually did the research and called the US Attorney General. The AG has not started to actively enforce this like spam, but hopefully it is coming.

Here is the federal statute that Click Fraud Violates. There are also a lot of state statutes as well. Click fraud is not only a civil offense. It is a criminal offence. If the Search Engine does not do anything to stop it or actually encourages it, they are liable. I got this directly from a DOJ attorney after I was on hold for an hour and a half. It is covered under the term "wire"

By the way, after I emailed this to the second tier company, I got a full refund that day.

UNITED STATES CODE ANNOTATEDTITLE 18. CRIMES AND CRIMINAL PROCEDUREPART I--CRIMESCHAPTER 63--MAIL FRAUD

§ 1343. Fraud by wire, radio, or television

Whoever, having devised or intending to devise any scheme or artifice to defraud, or for obtaining money or property by means of false or fraudulent pretenses, representations, or promises, transmits or causes to be transmitted by means of wire, radio, or television communication in interstate or foreign commerce, any writings, signs, signals, pictures, or sounds for the purpose of executing such scheme or artifice, shall be fined under this title or imprisoned not more than 20 years, or both. If the violation affects a financial institution, such person shall be fined not more than $1,000,000 or imprisoned not more than 30 years, or both.

http://www.usdoj.gov/criminal/cybercrime/1343NEW.htm

lots0

04-06-2005, 12:19 PM

...but also a zealous terrorist could do willful damage to a major Western capitalist industry in this way.You have got to be kidding! You better quit watching all that cable news, or at least filter the "so called" news with a little common sense...

DISC

04-06-2005, 12:29 PM

You have got to be kidding! You better quit watching all that cable news, or at least filter the "so called" news with a little common sense...

Actually, I agree that the terrorist threat is largely a fabrication. Griffin's eloquent and staggering book, The New Pearl Harbor, implies this, as do others. But this is for another forum.

lots0

04-06-2005, 01:10 PM

Here is the federal statute that Click Fraud Violates...
UNITED STATES CODE ANNOTATEDTITLE 18. CRIMES AND CRIMINAL PROCEDUREPART I--CRIMESCHAPTER 63--MAIL FRAUD
I agree fraud is against the law, but the people that brought the suit have to PROVE it was fraud and collusion.

As far as a criminal case goes, US history has shown that in most cases local and federal prosecutors will not go after US Corporations, especially for “white collar” crime.

I, Brian

04-06-2005, 01:52 PM

As far as a criminal case goes, US history has shown that in most cases local and federal prosecutors will not go after US Corporations, especially for “white collar” crime.
I'm under the impression that "Class Action Suit" is the real turner in this case, though.

krisval

04-07-2005, 11:20 AM

I'm under the impression that "Class Action Suit" is the real turner in this case, though.

I just want to clarify something for those unaware. The fraud law is dealt with in a "criminal" proceeding. The class action suit is in a "civil" proceeding.

The publicity of the Civil suit may just finally get the DOJ to start to take click fraud seriously and begin prosecuting against it. If you remember, it took the web community and specifically the ISPs and Portals to bring SPAM to the attention of the federal government before they finally created the Can-Spam act.

Bottom line. Although I hate lawsuits, I think this one will at least bring national attention to the issue and maybe clean up the industry. It needs it.

lots0

04-07-2005, 12:33 PM

I just want to clarify something for those unaware. The fraud law is dealt with in a "criminal" proceeding. The class action suit is in a "civil" proceeding.
Actually, it is not a class action suit yet, it has to be certified as a class action by the court, at this time all it is, is a simple law suit.

There is no guarantee that the court will grant class action status to this case.

fillossofer

04-08-2005, 04:12 PM

Yes, I believe news paper distribution is monitored - at least here they are :)

I worked at a newspaper for seven years and although paid circulation declined each and every year I was there "readership" increased to the point where the paper claimed you could reach 875,000 people by advertising in its pages. Our DMA consists of only 1.3M people! I always asked, how do you figure 875,000 people are reached when we only printed about 120,000 copies of the paper? The answer? People leave a paper on the bus, in waiting rooms, or copies are read by more than one person per household, etc. Yeah right, each copy being read by eight or more people!?! The official auditor of newspaper readership claims is the Audit Bureau of Circulation (ABC). Individual newspapers can conduct research on their own and have the ABC confirm their findings. Well guess what, the research firms understand that they are biting the hand that feeds them if results aren't in the best interest of the newspaper. You should see the questions on these surveys! In the last thirty days have you seen the headline from XYZ paper? OMG, I see the headline buying coffee in the morning, but that doesn't mean I read the paper! Anyway, glad to be out of that dying media.

Now having considerable experience with paid search I wanted to make a comment on the "conspiring" part of this suit. We placed PPC ads with about twelve engines over the last six months. First tier, second and third even. What is interesting is that in a few cases when researching the engines before placing ads with them, I noticed that our listings were already appearing in the results. Rarely the second or third tier engine would disclose in small faint type where the ads came from, and sometimes even the cost, but I started thinking - if I were paying this engine directly, would my ads appear twice? Would they override or eliminate the listing they were getting from the other source? Or worst case, count any clicks toward both campaigns???

What somewhat supported this theory was conversion rates. Where we'd see rates on first tier engines in the 13-15% range, second tier engines dropped off to around 6% and third to less than 1%. But, we'd be getting just as many gross clicks from all, supposedly from people typing in specific keywords. So, it would support the possibility that third tier sites we're charging us for clicks that were also being charged by another engine, and conversions were being attributed to the larger of the two. This is only a supposition as I have no evidence to support it, but it does make one think.

NFFC

04-08-2005, 04:15 PM

>Where we'd see rates on first tier engines in the 13-15% range, second tier engines dropped off to around 6% and third to less than 1%.

What was the cost per conversion rate?

Web Diversity

04-09-2005, 05:45 AM

This action will never wash and I think if Google, Yahoo and AJ roll over they may as well close the doors and leave the keys.

If you think about the way in which Google works out actual CPC it's based on CTR and includes impressions in the calculations. In some sectors SEO companies will run hundreds/thousands/millions of impressions running ranking checks and in doing so will artificially reduce the actual CTR as a result.

Clients do it themselves, I have had countless calls saying "i've checked on Google and my ad is in 7th place, so I looked again and again".

This is a case where the burden of proof will be on the people bringing the suit, which will be expensive and lengthy. Judges will struggle with the technology, after all, if we in the industry can't define what it is how can then judge on whether it is happening?

Within the ranks of the publishers there will always be those that fail to deliver results consistent with other publishers but does that mean the traffic is fraudulent? Or is it that their audience is not the right demographic for the keyword set?

As an agency we can ask for the removal of a publisher for the feed, and more often than not it will happen, providing the evidence is conclusive. If the supplier refuses then we have the right not to use that supplier.

At the end of the day if someone feels they are not getting value for money from paid search then they should stop doing it. It's new, it's fraught with issues of technology, misunderstanding, but inspite of all that it still delivers one of the best overall conversions of any other advertising medium, hence the reason it is growing at such a rate.

If I don't get any calls from a newspaper ad is that my fault or the fault of the paper? If they publish figures of readership in arriving at a rate card, that includes the free papers at hotels that probably end up being shredded or in a hamster cage.

If the court case goes ahead then this will be a huge victory for the networks, but it won't change the isolated instances when it will still go on. That is the nature of doing business online in 2005.

Webvisitor

04-09-2005, 06:37 PM

Intent will be hard to prove by plaintiff. Unless they can prove the companies in question set out to commit fraud they will lose.
The net result of this may be PPC providers and SEs having to clear out questionable affiliates. That would be good for all.

lazworld

04-10-2005, 09:06 AM

When Google, Overture, or other search engines give a client account a refund for click fraud they admit there is a problem but do not tell the public calculations used, or who was fraudulent, they issue a refund they deem appropriate, but it is accurate?

What are the parameters used to measure and detect fraud?

How many clicks were paid for those were fraudulent?

What is the average cost per click on fraudulent clicks?

People want answers.

James Liddell

04-10-2005, 07:50 PM

Is any of this actually illegal? I mean, is it against the Google T&C to do so? Do they promise to deliver only "human" clicks? And only so many from any one human in a day?

I know it is called "click fraud", but where is the fraud? is it the SE for not discovering the "fraudulent" click? Is it the clikck generator? Who?

As an analogy I have no expereince of, who actually audits the readership for newspapers and magazines? If I advertise because they have XYZ readership levels, does any independent group monitor this??

Besides which, surely the first port of call, short of a lawsuit, would be to ask for a refund. Wish there was a way to read more about the particulars of the suit... Anyone have any ideas on how to find an online copy of the suit??

Daily newspapers are audited independently by the Audit Bureau of Circulations. This is a fee-based membership that daily newspapers HAVE to subscribe to in order to appear in the semi-annual ABC Report which advertisers and advertising agencies use to measure cost/per/reader expenses. It used to be based in Ohio, I think. Weekly newspapers have a similar audit service. ABC pulls sneak attack inspections that can come at ANY time, day or night. The ABC team sent in has full legal access to the entire newspaper plant and offices and records The team starts by seizing the press run orders for that day and the three preceding days, observes the press run(s) of the day of arrival, checks the circulation records and subscriber lists to the Nth degree--literally a forensic audit, interviews field circulation managers and route delivery personnel and counts waste "ink-up" copies and press "clean up" ones. ABC claims accuracy to within 0.1 percent of actual sales. ABC is tough and extremely well-respected thoroughout the publishing industry. They also do magazines although mags generally use their own audit service.

James Liddell

04-10-2005, 08:13 PM

[QUOTE=fillossofer]I worked at a newspaper for seven years and although paid circulation declined each and every year I was there "readership" increased to the point where the paper claimed you could reach 875,000 people by advertising in its pages. Our DMA consists of only 1.3M people! I always asked, how do you figure 875,000 people are reached when we only printed about 120,000 copies of the paper? The answer? People leave a paper on the bus, in waiting rooms, or copies are read by more than one person per household, etc. Yeah right, each copy being read by eight or more people!?! The official auditor of newspaper readership claims is the Audit Bureau of Circulation (ABC). Individual newspapers can conduct research on their own and have the ABC confirm their findings. Well guess what, the research firms understand that they are biting the hand that feeds them if results aren't in the best interest of the newspaper. You should see the questions on these surveys! In the last thirty days have you seen the headline from XYZ paper? OMG, I see the headline buying coffee in the morning, but that doesn't mean I read the paper! Anyway, glad to be out of that dying media.

QUOTE]
I think you are referring to Reader/Per/Copy estimates. These are, indeed, estimates and are accepted as such by advertisers and ad agencies and are CLEARLY INDENTIFIED as estimates along with the factors used and the equations. Since most newspaper subscriptions are delivered to homes, the US Census average number of residents per household from the most recent census are used in the calculations. Most newspapers ethnically use residents above the age of ten since children very often read the comics and entertainment pages, including the specialized children's pages in weekend editions. There also are federal statistics from the, I think, Department of Commerce that are used in some way to estimate the number of readers per copy that is delivered to a business address. Newspaper circulation figures are the lifeblood of newspapers and is the only reliable thing that sets them apart from the Neilson Ratings for broadcast media, which are at best guessimates and are admitted to be such by senior commerical broadcast executives. I'm not at all familiar with how cable and satellite handle their viewer calculations but it must be similar to the Neilson system. I do know that they have tried to use some type of seeing eye device during sweeps which is supposed to estimate the number of people in a room with their eyes fixed on the TV screen but that is getting pretty exotic and I don't think it works very well.

Other comments on above post: I've heard about this reported fear by ABC of the "hands that feed them" and that has never been my experience. But then, I've not seen the entire spectrum of ABC. Eight readers per newspaper is unbelieveable high. I've never seen claims of more than 3.8 readers per newspaper copy and that was during the baby boom. Usually, the factor I've seen is about 2.3, sometimes as low as 1.8 for retirement communities. Your newspaper must have really been straining the system and those numbers would NEVER have stood muster with ABC or as a handout to an ad agency. Given to local advertisers only? I can believe that. As far as newspapers being a dying industry. Naw, they are generally holding their own overall particularly in the smaller cities and towns. The big problem right now for newspapers is reliable route people who will stick with it. The salvation will be the miniaturized, flexible display crystal and chip-PC embedded in a standard broadsheet plastic page to which each day's news and ads are downloaded from a coin-fed, counter-mounted device in restuarants or convenience stores, or at your home computer. No more circulation problems. Instant updates during the day. Take it with you flash video news too! Photocopy, electronic "clips" into a memory, all you can do with a regular newspaper. Cool.

projectphp

04-10-2005, 08:48 PM

James Liddell

04-10-2005, 09:34 PM

Thanx James. US centric, but still helpful :)

I always wondered how the ads in newspapers where measured. Good to hear there is some quality measure. I wonder if something similar will eventually be developed for PPC, some form of independant evaluation.

Ahhhh, I'm probably wrong--usually am in some details no matter what I do--but I thought that ABC was in most English-speaking countries. Hmmmm, could be wrong as that is only an assumption. Re PPC, here could be a business op for someone. For gosh sakes, I hope they don't go the Neilson approach. I interviewed a senior ABC VP for marketing once and he just plain laughed about the guessimates of Neilson AND his own marketing department. (We were off the record at that point). His rationale for using Neilson? "It's best thing we've got," he said. That's a heck of a recommendation for your prime yardstick, you know, especially since there isn't anything else. Newspapers are deadly serious when it comes to circulation numbers accuracy. There not a publisher in the industry who isn't fully aware that the solid numbers newspapers can put on the table are the ONLY reason journalists (I am very traditional and call electronic news people, newscasters or broadcasters. Huh, I don't know what I ought to call the news people on the Web--digress, sorry!) have been able to hold their own against first radio, then television , then cable, then satellite and, now, the Web. A dying industry? Sure have been a lot of opportunities to kill the newspapers but hasn't happened yet. In fact, they still have nearly all of their traditional market still in their hands--after 90 years of being told, you are about to go out of business, bud. There's something new in the market that is going to eat your lunch. --Ah, hasn't happened yet! :D

lots0

04-11-2005, 10:33 AM

Here is an article where googles Salar Kamangar director of product management defines click fraud.

defining click fraud, "or invalid clicks, as any method used to artificially and/or maliciously generate clicks or page impressions," ...

...Examples of invalid clicks, according to Kamangar, include: manual clicks on an ad to purposefully increase the ad spend; deliberate clicks on an ad to increase profits by site owners hosting the ads; and automated clicking tools, 'bots, or other deceptive software.
http://www.clickz.com/news/article.php/3453201

According to google they can detect click fraud, I wonder why they are not doing it?

, Google uses its own proprietary technology to analyze clicks to determine whether they fit a pattern of activity intended to artificially drive up an advertiser's costs.

"Our system automatically distinguishes between clicks generated through normal use by users, and clicks generated by unethical users and automated robots, enabling us to filter out invalid clicks," Kamangar said.
http://www.clickz.com/news/article.php/3453201

Ahhh the automated system to the rescue...

Maybe google should put some of that cheap 3rd world labor they hired to do some hand review...

<added>
there are already several companies that do 3rd party click fraud monitoring, just do a search for "click fraud"...

projectphp

04-11-2005, 10:50 PM

According to google they can detect click fraud, I wonder why they are not doing it?
Cost. Plain, simple cost.

Pretty much all methods of click fraud detection I can think of involve processing power. Some forms of invalid clicks require massive processing power, cross referencing multiple clicks, running processors over log files etc etc. Is that really feasible on every single campaign and every single click?

Nothing wrong with Google not spotting all click fraud either, IMHO. The cost of running servers to check for it will, inevitably, be pushed onto me as an advertiser / agency, most likely in the form of minimum bid increases.

So really, "stopping" click fraud may be bad for all concerned, and click fraud may be a risk to manage, rather than something that can be stamped out.

The question really, IMHO, is how much should SEs be required to do to find click fraud? How much should they be forced to find, stop and issue refunds for? The answer to that is not 100%, because that is impossible, and what we need is to find the "sweet spot" in which a portion of click fraud is killed, and refunds issued, and advertisers can choose to employ expensive (in terms of resources and processing power) methods themselves to uncover the rest based upon their own personal ROI expectations.

James Liddell

04-12-2005, 11:41 AM

Cost. Plain, simple cost.

The question really, IMHO, is how much should SEs be required to do to find click fraud? How much should they be forced to find, stop and issue refunds for? The answer to that is not 100%, because that is impossible, and what we need is to find the "sweet spot" in which a portion of click fraud is killed, and refunds issued, and advertisers can choose to employ expensive (in terms of resources and processing power) methods themselves to uncover the rest based upon their own personal ROI expectations.

Hi, all!
That "sweet spot"--a sort of optimalization point--could be defined statistically based on a true random sample of all "clicks" being used as the basis for billing an advertiser. Oh, I'm the guy who ponificated all over the map about ABC (Audit Bureau of Circirulations) for newspapers and Neilson for broadcast media a few days ago.

I can see the huge costs associated with a large-scale computer to calculate every single "click" to determine its either being fraud or honest injun. There really are only the two cases to consider you know. Gosh, the vastness of the Web and the number of advertisers and the number of users. Impossible to use a census. (That's from demography, you know--literally bean-counting everyone in a nation/state/county/city/tract.) BUT a survey (again, from demography--and used to inexpensively check up on the accuracy of a census or to update it in non-censusal years.) would be just the ticket.

Here's what you need to provide accurate and unchallenagable billing. One, a large-scale computer with a hidden address and secure against intrusion. Two, a staff of certified accountants and IT professionals. Three, a bonding company agreement. Four, a very accurate and generally accepted "viewing"-"tracing"-"recording"-and-"compiling" software. Five, clients--namely the search engine and browser companies, and even the advertisers--or an association to which the advertisers can choose to belong so that their voices are heard collectively. Six, an independent auditor of the measuring and reporting company I am describing here. And, Seven, publically accessable on-line periodic reports. Yes, that! No secrets, anywhere.

Because of the nature of the beast, eventually there will be only one Measuring/Reporting company. Monopolistic capitalism? Ah, here, yeah! The way to avoid this, perhaps, is to create niches for measuring and reporting service companies but that gets close to collusion and conspiracy and those are federal no-nos in anyone's book.

Having dealt with ABC for so many years and seeing the power of the audited and certifiable circulation numbers for newspapers as opposed to the Neilson guesstamations in getting ad bucks for the company treasury, I very strongly suggest that you people take the survey-bean-counting approach with independent auditing built-in. Yes, it will cost more but it will guarantee you as accurate billing information as is possible in an imperfect world.

I am absolutely certain that there is a very BIG business opportunity here for some whizkid. Go for it. It'll work. Just name it in my honor. That's all I ask.

Thank you for reading this.

Jim Liddell

lots0

04-12-2005, 12:08 PM

I can see the huge costs associated with a large-scale computer to calculate every single "click" to determine its either being fraud or honest injun. google uses lots of small cheap computers, not one or two big ones, that is one of the main things that makes google, google.

I agree that no SE is going to be able to completely get rid of CF. But, it appears that google is not even trying. How many advertisers have received a refund because of CF from google without asking for it? I have yet to hear of one.

I also agree there is a cost associated with detecting CF, but I don’t agree that is cost prohibitive.

Adwords & Adsense are google’s main source of income, spending a few dollars to detect click fraud to protect there business model is not unrealistic, in my opinion.

Do any of you know how long google operated before they made a profit? I think it was almost three years... If they can go three years without making any money, a few dollars to detect CF is not out of the question.

I do agree google not detecting CF has to do with money... refunding all that money would be hard to do, the stock might even drop if the word got out.

James Liddell

04-12-2005, 12:23 PM

What I keep hearing here are the companies being reported as doing their OWN certification of their billing claims. That, to my newspaper-trained mind, is unrelible by definition, and unsupportable under challenge. Internal auditing is always suspect.

What the industry needs to do is form a general trade association if it hasn't already and, acting in concern through the association, to establish or award startup competitive grants to would-be service industries designed to audit billing information. --Gosh, just as with the daily newspapers' Audit Bureau of Circulations. The way the newspaper so courageous set the ABC up left them as all blind men with the ABC having the only eye among them. Their creature was even encouraged to become the Frankenstein that everyone fears.

This Web Auditing Institute--to give it a name--would be totally independent of each and every individual company and their associations. It would be the 800-pound GO-rilla that each advertising medium can point to and tell an advertiser who is unhappy with a bill--"Go see them. We bill based on their audits. They are the industrial yardstick everyone uses."

Result, end of dispute. Who can argue with city hall? And the WAI would be "city hall".

But the key, critical, salient and most fundamental point I am making here is that the basis for billing MUST BE independent of the biller and the billee. It MUST be totally, completely and absolutely independent.

So far, based on what I've seen of the Web, Web personalities and Web stategies, it's still every-man-for-himself. Folks, that time is past, long past, and the sooner you start working together instead of taking delight in shooting each other, the sonner you will begin enjoy the harvest of what you have so amplied planted.

Jim Liddell

James Liddell

04-12-2005, 12:33 PM

[QUOTE=lots0] google uses lots of small cheap computers, not one or two big ones, that is one of the main things that makes google, google.

Ah, no longer so, or soon to be no longer so. Google recently accquired a monster complex near Douglasville GA to use just as a computer base. It is supposed to be ultra-secret but a local newspaper spilled the goo-goo-googly beans and even ran a picture of the place. (well, a picture of one corner. The building is too big to get into one frame.)

Hope it isn't jinxed, for Google's sake (Well, I don't know about that, actually. Google just partnered with Al Gore to bring out a kid's political discussion TV channel--and I didn't vote for Big Al.) The building was built for a dot-com busted bubble outfit that didn't even get to move in before they bellied up in the early days of the collapse.)

lots0

04-12-2005, 12:46 PM

Google recently accquired a monster complex near Douglasville GA to use just as a computer base. It takes a lot of space for all those small cheap networked computers...

All the exact locations of the google data centers are "secret". Wanna bet the newspaper that broke that story just cost their community a new business...

Google can't afford for security reasons to let the "public" know the physical location of their data centers.

I'll bet Matt C. was fit to be tied when that hit the papers.... LOL

James Liddell

04-12-2005, 07:02 PM

It takes a lot of space for all those small cheap networked computers...

All the exact locations of the google data centers are "secret". Wanna bet the newspaper that broke that story just cost their community a new business...

Google can't afford for security reasons to let the "public" know the physical location of their data centers.

I'll bet Matt C. was fit to be tied when that hit the papers.... LOL
The story deliberately did not mention the address and unless you are an local insider you probably won't know. The place is mostly a bedroom community except for regional retail anyhow. Thanks for insight on other secrets of Google.

hardball

04-13-2005, 09:52 AM

PPC is simply a road tax for web players, there should be a demand for at least fair taxation. The argument that PPC and internet infrastructure don't play well together as it relates to fraud is preposterous, would you blame the road for occupant fallout if you manufactured a car without doors? The engines are probably grinning like a cheshire cat with this bit of illusion.

If PPC doesn't work on this road then PPC is simply flawed, many other transparent models are available.

MikeDammann

04-17-2005, 07:48 AM

I just came accross this Washington Post article
http://www.msnbc.msn.com/id/7529767/
$50k in click fraud by only one company.

Mikkel deMib Svendsen

04-17-2005, 08:57 AM

The above mentioned article is yet another blatant example of how search engines reject the seriousness of click fraud. When Google's Kamangar say: "Google's technology detects invalid clicks..." most of us know that is simply not very true. They do detect some invalid clicks but far from all as several cases have shown. Also, as Jessie points out in the article it can be pretty difficult to get money refunded even if you do have good data to prove your case. And it shoudn't have to be so.

I think search engines need to provide a third party validation the same way printed media and TV does in some countries.

This is not getting better. In fact, I believe we have just seen the beginning. There are so many reasons people will initiate click fraud (not even a fraction of them covered in the article) and so many ways to gain or profit from it - or simply just create terror (for fun or "political" reasons).

I don't think this will just gradually get worse - or better. I think we will sooner or later see some kind, or kinds, of major attack(s). There are already hackers and virus makers that have control of millions of computers today that could be used for very effective and almost impossible to detect attacks. The big question is: Are the search engines ready for this? I think not! And this is where the real bomb under search as we know it lies. None of us, I assume, really want that bomb to blow up.

Off course, by time the search engines could probably stop such a bomb but would you keep throwing your money into the pool after a click fraud bomb goes off untill they find a "cure"? I don't think I would.

OK, I know that is a very paranoid vision :) However, it could very well become real unless the engines start gearing up a serious way on these issues. I really hope they do ...

James Liddell

04-17-2005, 10:42 AM

Mikkel deMib Svendsen

04-17-2005, 10:59 AM

I think it will take a great deal of preassure from the advertisers to get the search engines to spend the money it will take to get a proper third party validation. So let's begin ... :)

In fact, this is one of the reasons we have been seeing such a great interest in an organisation such as SMA: SEMs want to unite forces to improve this (among other things). And trust me, if it's up to me the engines better get ready for some SERIOUS pressure - at least as long as I am part of SMA

Having said that I don't think any third party will be able to do much about the effects of a real click bomb attack - they can only report it ... "today we got 95% fraudelent clicks" - and what will that do? However, a third party will be good in validating weather it's the engines that are right when they say "this is not a problem" or experts such as Jessie when she claim it is. It will legitimize our business and that should be good for all - engines and SEMs alike. I wish the engines would understand that.

projectphp

04-17-2005, 11:51 AM

I think it will take a great deal of preassure from the advertisers to get the search engines to spend the money it will take to get a proper third party validation.
Mikkel, before we all go doing that, can we decide what it is first that we actually want? Do we want SEs to stop click fraud, or do we really just want more transparency, like standardised forms to ask for a refund, FAQs on what information to include for a refund etc.

And can we understand the consequences of our choices? I mean, great, Search Engines start paying for click fraud detection. What is the catch? Minimum bid increases? Slower in delivering on other enhancements... what?

As I see it, there are many types of click fraud, some of which can be stopped, some of which can't, and some of which can be deterred with simple functionality enhancements.

As an example of the latter, http://www.lurhq.com/ppc-hijack.html is a great article, but what it outlines isn't really PPC click fraud. What it outlines is DNS hijacking to generate PPC $$. So the fraud isn;t the click itself, but perpetuated elsewhere.

Now, there are several ways this example could be stopped, one of which is to have greater control over exactly which sites an ad is shown on.

If one could choose, for example, Google, AOL, Netscape and Ask Jeeves and nothing else for AdWords, that would effectively stop some of the smaller affiliates from utilising click fraud for their own benefit.

Getting this sort of functionality enhancement on PPCs would be a double win, as it will help curb some click fraud and there are certainly other uses for it.

I know that arguing that click fraud is something we should think about before trying to get Search engines to pay for it is an unpopular view. I just worry that stopping it may cost PPC advertisers more than allowing it to continue, especially if it is applied to every account when so few accounts probably notice or care.

Mikkel deMib Svendsen

04-17-2005, 12:11 PM

I just worry that stopping it may cost PPC advertisers more than allowing it to continue

I don't think so. I believe the third party monitoring is one of several things we, as advertisers and SEMs, should ask for. Another thing is the ability to gain better control of where our ads are showing. Thats what I think and thats what I will fight for. Very soon we will have elections for SMA and then it'll be up to the members to decide if they support me or other views. If enough people think like me I believe we an make an impact and possible change things. And thats what I think we should do.

PhilC

04-17-2005, 12:55 PM

I think we will sooner or later see some kind, or kinds, of major attack(s). It's already happened to Google. Perhaps not on the massive scale that you described but it happened on large scale about a year ago. On that occassion they knew it was happening, but they paid out the affiliates in full. I can only assume that they also charged the advertisers, but I don't know about that.

Web Diversity

04-17-2005, 08:13 PM

The view on click fraud is always heavily stacked against the search engines providing the service.

If you look at a service like eBay they merely provide the means by which buyers and sellers get together. The community itself assesses the worthiness of the sellers ability to fulfil.

As an agency we have rarely used the content network, primarily because in the case of Google there is no qualification criteria in having adsense served up on a site, and whilst the fact they don't tell publishers what they are getting for the clicks it still encourages the sort of behaviour that loosely constitutes click fraud.

In some instances where we have opted in to content network traffic the results have been pretty good, but there are still some rogues in amongst the good.

It's only by working together with the engines that we can reduce the instance of rogues being able to benefit from ill gotten gains, but in order to do this advertisers need to have the right tools in place.

If you placed the cost of click fraud up against the cost of say usability flaws, site downtime, inappropriate ad creation by novice advertisers then click fraud is a pimple.

I always smile when I read about advertisers claiming their competitors are clicking on their ads. It is blatantly obvious that this will go on, every phone call we receive from a client who is looking at the performance of their own campaigns will say things that lead us to believe they must have clicked on their competitors ads to know what they are up to. It's part of doing business online.

If advertisers are losing money on PPC then they should stop and try something else. You rarely hear the good side of the services that Google, Yahoo etc. provide. The cost of click fraud may be a lot, but a lot of advertisers have made significant profit on the back of cheap advertising, let's not forget that.

If Google/Yahoo opened up the floodgates for advertisers who suspected they had been victims then they would both need twice as many staff just to cover those off. They'd spend all day looking at logs for $2 refunds, which will cost $50 to investigate.

Fire prevention is much better than fire fighting, we should be looking at ways to out the rogues, not to kick the suppliers for being cagey and protective on how they police it internally.

We've worked with technicians on all major PPC's and if they don't care less about click fraud then they do a very good job of cloaking their feelings, but like anything the burden of proof is with us.

Mikkel deMib Svendsen

04-17-2005, 08:26 PM

The only thing I agree to is that it is a business issue. However, just because my clients profit from PPC today dosn't make it right that they have to accept any fraud. They could probably profit more if the engines would put more resources into click fraud detection, filtering and better handling of reported fraud. Who have the right to say when my clients profit enough? Certainly not the engines. Sorry, but that is just not a very good response in my opinion.

The other side of it is that a lot of companies I deal with do in fact leave PPC and move more and more money into organic because CPCs are getting too high and conversions too low. In many competitive industries I work I can produce more organic visitors than PPC-visitors for my clients and at a fraction of the price (one recent case turned out to be approx $1 CPC for PPC and $0.20 for organic - with twice the volume). So, I do actually already see the effect of degrading quality. I cannot, off course, be certain what part if the degrading quality that comes from click fraud and what is not but I certainly see the move happening.

Web Diversity

04-17-2005, 09:02 PM

If clients want to make more profit then of course moving some or all of their budget to organic is likely to yield better ROI and the value of a good SEO is priceless.

But if advertisers want to compete in the cut-throat world of online advertising, which, as we all know is one of the fastest growth sectors bar none, then they have to turn up and be ready for what they might face.

Policing any sort of questionable activity costs money. Yes, some advertisers are more prone to suffer due to their sector. These are the ones that will benefit the most from employing the best around to help in their quest.

I am sure if you looked at the sectors concerned many of these are the ones that have the biggest potential gains. Publishers most likely to perpertate click fraud will do so in arenas where the gains all round are likely to be the highest. The methods adopted are almost unimportant, it's the intent that is the problem.

It might sound glib, but getting the money back is the easy part. It costs time and human resource to investigate and often the cost of the investigation will run to more than the amount involved, so what should you do in that case?

If the suppliers don't do more to prevent click fraud then advertisers will vote with their wallets, revenues will drop, share prices will fall, confidence will wane.

I'm with you on lobbying for them to do more, but I don't see it just being a "chuck a pile of people or money at the problem and expect it to get better" solution.

The more money is spent online, the more there will be a desire for ad inventory, the more this problem is going to continue.

Good luck with the elections Mikkel, we need good lobbyists!

Mikkel deMib Svendsen

04-17-2005, 09:11 PM

Good luck with the elections Mikkel, we need good lobbyists!

Thanks :)

Just to make myself clear, I don't think there is one great solution to all problems and you are right that all the money and people in the world won't magically make all click fraud go away. But if more resources are put to it I do believe it will get better and it will potentially help create stronger protection shields against serious future click bombs. I am not sure what the right level of protection is but I strongl feel it's not high enough as it is. I have also seen too much evidence that search engines often fail to act correct in these matters. It has to be better. Better is not perfect but it's at least better :)

MikeDammann

04-18-2005, 12:40 AM

What worries me the most is not the current cheaters but the ones coming up daily. All they need is a free blog and they are in.
I have seen another thread where someone bragged about getting people from chat rooms to click on his ads.
This seems to be getting worse.
I don't have a solution myself but like Mikkel said, raising awareness is definitely important.

Web Diversity

04-18-2005, 06:02 AM

More resource is definately the way forward, and as an agency we have allocated a lot more resource to identifying and eradicating the fraudulent traffic because ultimately it has a detrimental impact on the overall ROI.

That being said though it has meant a different view in terms of charging structures as it costs money to save money. All of us will be playing catch up with the perpetrators and it may be that some sort of legal fights will make everyone sit up and take note, but where there is greed there will always be opportunists.

Mikkel deMib Svendsen

04-18-2005, 06:12 AM

I actually don't think that it will cost the advertisers more if the engines spend more money to protect their systems. It' dosn't work like that. Advertisers don't id our of pitty for the "poor engines" - they bid based on what makes sense to them. I won't bid one cent more just because the engines stat facing up to the problems they have with their own media. No way! I will continue to bid, and advice clients to bid, what makes sense to them. And I am pretty sure the market will follow - as they always has. You never pay for advertising based on production costs of the medie but on the value it gives.

If protecting the system, by the end of the day, cost the engines more than they profit then PPC may not survive, but I doubt that will happen. Look at the profits they are pulling in now. I think there should be room for quite a few more people on the case without going broke :)

MikeDammann

04-18-2005, 06:22 AM

Somebody is paying right now and not getting what they pay for. That is the advertisers. Saving them money would be best for everyone. Increase their ROI and the overall income for adsense affiliates will rise.
The ppc amounts will go up and so the income because the clicks overall generally speaking will increase in value. Same way Google and the advertisers, everyone should benefit.
The question is how to prevent click fraud and what kind of algorithm it would take to catch it.
IMO the improvements should be made in who to accept and who to decline as an affiliate.

Web Diversity

04-18-2005, 06:58 AM

Site Tutor you are spot on there.

With confidence in the system advertisers spend more, so everyone should be a winner.

We don't isolate questionable traffic in a pigeon hole on it's own. The impact of a site being down and the clicks being delivered will be just as damaging as a fraudlent click so we try to work on ways in which ROI can be improved and monitoring the source and outcome of every visitor is an important part of that process.

If you identify what normal behaviour looks like then it's easy to spot the stuff that is deviating from that and identify what can be done to improve it.

James Liddell

04-18-2005, 10:37 AM

:D Yeah, I'll take credit for the direction these exchanges are going in!!

Here's a few observations to keep them focused.

1. The pressure on the search engines will begun self-generating. Large advertisers will start belly-aching about the accuracy of their billing. In other words, the marketplace will develop a momentum in the direction you want.

1a. A few well-placed lawsuits by large advertisers will also force the search engines to find a better way to keep hostile L-A-W-Y-E-R-S out of their lives.

1b. Small advertisers will create trade associations, if they haven't already, to complain, whine and belly-ache.

2. I see a consuses forming that a third-party auditing/certification/reporting service TPACRS) is a good way to answer all this. That idea needs to remain at the heart of all this, NO MATTER what final shape it is given or how it comes into being.

3. The cost of TPACRS will be less than one-percent of gross revenues for ANYONE AND EVERYONE! Why? Because TPACRS will be a one-shop replacement for all the presently duplicated services across the industry and will eliminate all that wasteful douplication. And because TPACRS will oversee the authenication of EVERYONE's billing-"advertising medium" and "advertisers" alike, it will begin developing efficiencies of operation (improved methodology), economies of scale and legal order to the system. After all, how expensive to the industry is that company that maintains domain-names for everyone, huh?

4. As far as my remark that a strong TPACRS will reduce the hackers' effect. Well, it WILL, by definition. I firmly believe that! --Here's why. Since it would be the SOLE click-authenicating entity for everyone, the vulnerabilities of the marketplace overall would be reduced--because its staff would know what Google's self-defense force knew AND Yahoo's AND Ask Jeeves and so forth PLUS developing and adding its own defenses based on that amagamation of industry-wide knowledge provided to it by the "advertising media" of the Web.(That's all the search engines and related are, you know. The Web's "advertising media"). --Gosh, I sure hope I'm not tossing out too many new perspectives and concepts at a time here.

And you cannot tell me that Yahoo's folks know everything that Google's folks do regarding defending against click fraud and hacker attacks, and visa versa. In discussing this, I firmly maintain that it is far easier to defend ONE computer system (TPACRS's) against intrusion than two dozen (Goggle, Yahoo, etc.).

OK, that's one side of the problem you guys are looking at. The other is fraudulent click manipulation by the hackers--where they do their mischief out in the field where the clicks are being created. Well, I said at the outset that TPACRS would have to use statistical sampling to certify the number of legitimate clicks that are used to underpin the billing to advertisers. I think everyone will agree that NO ONE can count the universe. So, statistical analysis is the only way out of the thicket. And, luckily, statistics has in recent years become, today, a most remarkable and reliable tool. Statistics can be used to discount hacker influence through ANY NUMBER of statistical procedures--surveys and cluster analyses are just two of dozens of tools used by statisticians. Will the final reported figures be perfect? No, this is an imperfect world--just as it IS POSSIBLE for a really savvy newspaper circulation manager to "goose" his numbers even under the eyes of the Audit Bureau of Circulations. But, you know what? Heshe can't do a lot of it and he/she won't be able to do it for more than, say, two years before having to stop or risk being caught red-handed.

But this is the very point I keep returning to. TPACRS will be the "best thing we have" (and will be pretty dang accurate anyhow) and will be the umpire that everyone will come to accept as the honest broker everyone can appeal to and get a fair deal from in return.

Gosh, I sure do write long messages, don't I. I gues its that bag of unused words under my desk that was left over from my career in journalism. I just hate to see good words going to wastel, so you guys end up having to read them.

Luckily, I think I've run out of things to say.

Jim Liddell :) :rolleyes:

andrewgoodman

04-20-2005, 05:16 PM

:) You raise new and valid points about concern over the validity of accurate click counting for ad-billing purposes AND the threat of hackers to interfere with the self-monitoring/self-auditing approach.

You also touch on a third-party verification proposal which I discussed at length in my message #36 a few days ago. I invite you to scroll back a page on the string and read it.

The ONLY possible solution lies not in self-monitoring/self-auditing but in the established and subscription to a third-party click auditing-and-reporting service--one with sufficient resources to continually outfox the hackers and one with sufficient clout, and industrial authority and respect to make everyone 'toe the line' or risk being dropped from the system, and thus, lose billing authenication.

Trust by bill-payers can ONLY be earned and retained by abandoning self-audits--which are ALWAYS suspect and rarely acceptable in courts of law--and accepting the services of a totally independent >>third-party<< service.

Might this involve revealing some corporate secrets so that the service can do its job in the most efficient manner? Yes, very likely--particularly in the design of the search engines, themselves, and in the in-house click counting and certification systems as well, but "secrets-sharing" things like this are done every day by means of trade agreements, nondisclosure certificates, and contractual arrangements. There would be nothing new here, legally, at all. Everything would be "off the shelf".

Again, I tell the world that there is a HUGE business opportunity here for some go-getter. Just "ape" the newspapers' Audit Bureau of Circulations. There, a tried and true, proven blueprint is already in place and ready to be copied for the use of the digital world. Huh!--Here's a thought! Why can't the ABC be asked about its interest in becoming this service? :)
This is a story with a fair bit of history, ladies & gents. Shortly after I wrote this 2001 article (http://www.traffick.com/story/2001-08/abcinteractiveaud.asp) about the Audit Bureau of Circulations Interactive unit, the unit was folded back into the parent company / downsized.

Obviously, the model is there.

In an earlier generation of web advertising (under siege, undergoing a crisis of confidence), a few companies turned to ABCi to audit traffic patterns, but it seems to me it was more of a dispute resolution approach. Few had much stomach for it. Obviously if the market had been there for it, ABCi would be going gangbusters. No one was buying it so they stopped selling it. The last news release by the unit was on August 5, 2003.

In any case, now that we have seen a changing of the guard, to the point where traditional online ad brokers like Doubleclick have faded, and Yahoo and Google dominate, it is time for these new leaders to step up to the plate and advocate for the creation of some third-party-supervised standards a la Audit Bureau of Circulation.

But it is a chicken-and-egg question. The leading ad players need to help pay for the service. The service needs to be there and the Audit Bureau needs to be ready and willing to provide it. I don't know how it should all work but I assume there needs to be some institutional push through organizations like the IAB to bring different stakeholders together to figure out how to set up an independent auditing standard for various forms of online advertising.... be they pay-per-click or anything else.

I don't think it will fly if it is just some opportunist building something from scratch and hoping it will catch on. Everyone needs to agree on the model, and it would help if it were the Audit Bureau itself that built or rebuilt the division that dealt with online advertising auditing.

James Liddell

04-20-2005, 07:39 PM

:) Goodman wrote a good and apparently insightful article and the posting I'm commenting on just above has as well. I was totaly unaware of ABC's past involvement and it pleases me that nearly all my points have already been addressed. I just stumbled into this discussion and felt I had something worthwhile to add. That task is done and I'll be interested to see what further developments come along.

In summary, third-party auditing, a sincerely felt industrywide need for a new approach to certifying the ad bills, and statistical sampling used as the basis for that--the three elements needed to solve this issue so everyone can get back to more mundane business.

Nice discussion. Thank you for letting me have a voice in it.

Jim Liddell :D

hiero

04-27-2005, 02:38 PM

About 4 or 5 years ago I approached Overture, who may have been Goto at the time not sure I forgot when they switched over with a what I thought was a click fraud problem. My ad account averaged about 100 click throughs a day and I was spending around $50 per day. In a 3 month period I would see spikes in traffic that would send my daily price to $500, and my order rate would get worse and less. To me this made no sense, if I was getting more traffic why wasn't my conversion rate improving also. To make a long story short after going around and around for weeks they stated that the charges were accurate due to an increase in searches for that particular term. I never believed it.

I knew this lawsuit would come, it was just a matter of when.

James Liddell

04-27-2005, 03:59 PM

I just came accross this Washington Post article
http://www.msnbc.msn.com/id/7529767/
$50k in click fraud by only one company.

Regretfully, the link has been broken. This happens often with newspapers with limited computer power to keep stories up. I just wish people would go to the trouble of a copy-paste when using ANY newspaper or magazine. Now, none of us will know what was in that story. Well, one guy did because that's why I came back to read about some lawsuit. Gee. (Hmmmm, no sad smiley face. So I'll make up one. (Sad smiley face?!) :-(

James Liddell

04-27-2005, 04:20 PM

About 4 or 5 years ago I approached Overture, who may have been Goto at the time not sure I forgot when they switched over with a what I thought was a click fraud problem. ... was spending around $50 per day. In a 3 month period I would see spikes in traffic that would send my daily price to $500, and my order rate would get worse and less. ...they stated that the charges were accurate due to an increase in searches for that particular term. I knew this lawsuit would come, it was just a matter of when.

If you kept the details on this lawsuit, I 'd like to see them. But my main thought here is that what you are doing is handing a signed and dated blank check to your advertising medium and in the doing telling them to "gouge me good". I could use a different word for "guoge" but there could be a censor here and the word is rather impolite as well. :p

Jim

hiero

04-27-2005, 04:43 PM

James Liddell

04-27-2005, 08:13 PM

James I still have the email and the excel spreadsheet that they sent me showing all of the recorded traffic to the site. I can email it to you if you want to look at it. It was an ad campaign for one of my clients. Ever since that day I have spent much less time on PPC campaigns.

Hi,
No, I would not learn a thing from reading it because I don't have the context for the Excel even with the email. What I was interested in are the details in the news story you cited. No big deal. No need to write out a long msg about it. I accept that there is/was a major lawsuit over click count and billing.

Your shifting your client's ad placement is exactly what is going to increasingly happen until the "ad media" (search engines, etc.) catch on that they are only injuring themselves with in-house billing audits.

With established news media and other ad presenters, you know the rates, you know the placement and you know what the final bill will be because you write the bill when you send the insertion order for your ads. I know you know this but I'm stating these facts here for the benefit of the digital crowd who probably have never placed an ad in a newspaper, or with a tv, magazine or radio station--or with a billboard company, or direct mail house, etc.

In fact, this digital world's billing based on clicks is the only major application I know that are somehow related to production of responses for an advertiser. The only other applications of this type that I can think of are the early morning infomercials for salad cutters and exercise equipment and investment advice books--which is a pretty sorry gang for a major ad medium to fall into. With these cruddy outfits, as you well know, the tV station is reimbursed on the basis of how many sales it generates for the whatever peddlers. Huh! But even here, the payment is based on actual purchases and not on number of "views" as it were as represented by the number of "clicks" which are not the same as number of purchases.

Myself, in your position, I would take the position with my clients of "Yeah, probably nice but no control over ad costs. Let's do newspaper inserts or a direct mail campaign using tested mailing lists from a reputable list house."

Enough, again, from me on this subject. My position is quite clear.

Jim :D

hiero

04-27-2005, 10:02 PM

You nailed it on the head Jim http://forums.searchenginewatch.com/images/icons/icon7.gif

Meanwhile the PPC market continues to grow by leaps and bounds and hopefully the fraud has become less and tracking has become better.

James Liddell

04-28-2005, 04:01 AM

That is accusatory of some act that usually results in a major prison-time-like penalty! People don't like to be charged with that. Without evidence, it rebounds as libel. OOPS! And that is expensive if your shyster is a loser.

Beside, there is also the hackers (game-players) who will just love to introduce havoc to this area and I think that will be an even bigger problem.

I don't see long-term happiness until third-party auditing, etc. as I've too verbosely already discussed.

Thanks for the kudo. I really have nothing more of substance to add to this discussion. Bye.

Jim :D

Lori Weiman

07-13-2005, 01:42 PM

Lots of lawsuits trying to form class actions - anybody think advertisers will join any of these? We track clicks and measure fraud, and so far, none of our clients seem interested. Anybody have an opinion on this?

PhilC

07-13-2005, 02:02 PM

It doesn't surprise me that they don't seem interested. They probably realise that it won't succeed. No court is going to find anyone guilty of missing some fraudulent clicks. A court is going to find that it's much the same as shoplifting - it's to be expected.

The problem with a case like that is that it has to be shown that Google isn't doing enough to prevent it, and, unless there is real evidence, it can't be shown.

However, I'm interested to know what makes you think you can measure click fraud. If you see clicks from the same IP coming too quickly, then you can measure it, and Google should also have seen it and not charged for the clicks. I'd be very surprised if that kind of thing gets through and is charged for. If it does, then they really should be taken to court. But there's not much else you can do to measure it. It's very difficult to say categorically that some clicks are fraudulent and some are not.

Lori Weiman

07-13-2005, 02:32 PM

you're right, its impossible to determine the intent of a click and to measure click fraud definitively.

We look at risk factors and showcase what is believed to be suspicious - meaning worthy of further investigation and/or likely to be clicks committed with the intent to defraud the advertiser. The evidence is circumstantial always. There are many factors: successive clicks in the same day from the same IP on the same word - which will be caught by the engines if they are in close proximity to one another but will not be caught if there is a time delay of x minutes (no one knows for sure the formula deployed by the engines to determine what "successive" means), succesive clicks in the same day from the same IP on different words (we suspect this is not caught by the engines), abnormal click spikes from one affiliate or one IP with abnormally low conversion rates when compared against other IP's and affiliates, international clicks from places where the advertiser does not choose to sell products, and clicks received where there is no referrer. All of these are factors, but are not necessarily fraud.

Still curious - anyone thinking about joining the current class actions?