I recently came across this article, thanks to a link from a SearchEngineWatch newsletter:

http://www.infosyssec.com/forum/viewtopic.php?t=1496

This author is arguing that 'click for dollars' scam programs are generating fat paychecks for scam webmasters while wasting LOTS of PPC advertising budgets. This could happen by Mr. Scam Webmaster signing up to have PPC ads displayed on his site. Then, he hires folks to repeatedly clicks those ads. :mad:

I'd LOVE to hear some thoughts on this possibility. I checked the two PPC campaigns I'm being paid to manage, and a large chunk of their budget is going to content-distributed sources, i.e. PPC ads being shown on other sites outside normal search engine sites. SO, my clients could be falling prey to this attack.

Thoughts? Comments?

I recently came across this article, thanks to a link from a SearchEngineWatch newsletter:

http://www.infosyssec.com/forum/viewtopic.php?t=1496

This author is arguing that 'click for dollars' scam programs are generating fat paychecks for scam webmasters while wasting LOTS of PPC advertising budgets. This could happen by Mr. Scam Webmaster signing up to have PPC ads displayed on his site. Then, he hires folks to repeatedly clicks those ads. :mad:

I'd LOVE to hear some thoughts on this possibility. I checked the two PPC campaigns I'm being paid to manage, and a large chunk of their budget is going to content-distributed sources, i.e. PPC ads being shown on other sites outside normal search engine sites. SO, my clients could be falling prey to this attack.

Thoughts? Comments?

Okay first thing go inside your campaigns and turn off the content search portion is one solution... content has not worked well for many people.
Second get a fraud tracking program.. there are many out there such as at KeywordMax.

Thanks, AussieWebmaster - I already deselected the Content Network.

whos clicking who is another fraud tool I have heard of

if you are not on content networks then people are not clicking your ads to generate revenue. competitors may click them to cost you money, but even when this happens it is partially offset by cheaper click price since the effective clickthrough rate has gone up

I would also like to point out that tracking conversion can be a good indicator of when something isnt right. Not too mention if you have no/low conversion, and your not branding. Thats a good indication to either rewrite your ads or dump those keyword buys all together.

Thanks, AussieWebmaster - I already deselected the Content Network.
So you had this deselected at the time of the problem? You need to look at the IP addresses of the clicks... if there is a large number of duplicates it could be a competitor looking to waste your money and get you to your budget limits so they can hold the spot for less.

I have a few comments.

- On fraud detection tools, I don't see what they offer over and above the more generic category of analytics / user tracking software. Most analytics packages offer a range of features and certainly anyone using one they're comfortable with and using it to the fullest can glean raw info about potential fraud there, and then use it to report potential violations to Google.

- I don't agree that turning content targeting off is the answer. Turning it off is a decent "when in doubt" answer, but in practice many advertisers are in expensive keyword areas and are also seeking more volume, and *if bid lower*, content targeting can and often is effective. There are ways to bid lower on it to see if it performs for you at a lower average CPC, they're just not widely used or understood.

- The real problem, with Google and Overture both, is that there is not enough channel control over content targeting. Meaning, you cannot even shut off the "bad apples" in the absence of Google/Overture taking action to police them. The technology is obviously there -- in AdSense, for example, you can type URL's into a box as a publisher, and no ads for those URL's will show up on your site. So similarly, as an advertiser, why not be given the ability to block your ad from appearing on sites that seem to be draining your funds by submitting a similar list of "blocked URL's" into the AdWords interface? And here's a neat idea... why not then make an aggregate "most banned sites" list available to advertisers so they can be aware of potential abusers? At the very least, even if not made public, the "most banned sites" list would quickly come to the attention of Google, so they would have a more rational way of policing potential abusers without having to rely on anecdotal evidence and semantic arguments from email and phone customer service exchanges with upset advertisers.

The fraud question is complex and we have to be careful not to paint PPC with a broad brush given the ROI it offers compared to traditional marketing channels. We do have the ability to overcome such problems with creativity mixed with data, and within the boundaries of the marketplace. As long as we are given some better tools to stand up for ourselves as advertisers vs. unscrupulous or simply poor-converting publishers.

Google would never want to publish a *most bad apples in OUR PROGRAM* site, but whoever is the smart marketer who makes one first will likely get a ton of free link popularity from within the seo field.

that is almost a fun enough idea to make me want to learn PHP... :)

Andrew I agree it can be done with due diligence of log files... but advanced tracking makes it quick and easy... the same goes for the fraud tracking... I am not down on the content search per se, all elements of the internet can be hacked...
I find that search is a better fit and has a better close ratio for me... content can work but it takes a lot of work and there is a bit of a problem when search terms have to be in search to be in content and then you want a good position for search but maybe a lower one for content... I am actually spending some time and money right now trying to solve the puzzle... if I get somewhere with it I will post info here.

I have been vocal to the people at Google about the publication option that places like Blow Search offer. It may come eventually, but not any time soon.

In the mean time a list would give Google something to look at and am sure they would... though the possibility of libel etc. would make it a fine line to walk.

andrewgoodman - that is a great idea of allowing advertisers to opt out of certain sites. However, I don't know if it could really solve the problem. I'm guessing those scam webmasters are pretty slick and could easily move their operations to different domain names in a hurry.

Don't get me wrong: I'm NOT knocking PPC advertising. I think it offers potentially fantastic ROI and value for money. Plus, you don't have to convince customers to try your product or service; they are already actively seeking you!

Another plus is that PPC is a quick fix to SEO issues. I can immediately get on Google's first page when my actual site may rank on page 4,000.

Furthermore, PPC allows Mom and Pop to play with the big boys. I've got one campaign for a small company that usually has their ad ranked #1. Find targeted enough keywords and bingo, bring on the qualified traffic.

All I'm saying is that PART of the system may be flawed, and I would want to warn others of that. I personally have opted out of Content Search.

Not necessarily -- the "scammers" can't use a new domain name, because they'd need to apply and go through an editorial process. Now is the time for Google to make the editorial process more stringent, and stop letting in so many new ones so quickly. The existing base needs to be cleaned up first before they expand.

This used to be a bigger problem for the very reason you state: when GoTo was distributing its listings through low-quality affiliates back in the day, they had limited resources to police new applicants. Now with $billions in the coffers, Overture and Google have the resources to be more watchful of affiliate publishers.

That doesn't mean some bad apples won't sneak in, but if they do so repeatedly, their names and other identifying factors will be associated with the fraud, and I wouldn't want to be in their shoes once someone decides to sue or prosecute.

(The law does catch up. On a different topic, the "father and son spam team" from Kitchener Ontario have paid Yahoo a six-figure settlement and have signed various documents vowing to go out of the business and vowing to go around educating people about how bad spam is. I guess that's a good way to stay out of jail.)

It will also get harder for fraudulent publishers to keep at it, I feel, but along with our own tracking data, we need the PPC players to assist in blocking the bad guys.

I guess my content advertising isn't costing me that much, because I am not even slightly worried about someone in Mumbai happily clicking my listings all night long. In fact, I am almost certain I'm getting a little of that because one of my ads is showing up on a bunch of these "nigritude ultramarine" pages.

Even if I were worried, getting the ability to "opt out" of certain sites wouldn't do me any good anyway, since I would have far less of a clue than Google or Overture about which are the bad ones.

OTOH, I see content advertising as a branding thing anyway. The CTR is so low (like 0.2% or whatever) that it becomes very low cost advertising. When the CPM is 25 cents, you tend to overlook a lot of other factors.

Much more useful would be the ability to create separate campaigns and ads for content-based advertising, so you can control bids, and target different keywords. Let me have a checkbox so the ads *don't* show on Google.com, but only on Adsense. I'd put a bigger budget on that than search advertising.

I guess my content advertising isn't costing me that much, because I am not even slightly worried about someone in Mumbai happily clicking my listings all night long. In fact, I am almost certain I'm getting a little of that because one of my ads is showing up on a bunch of these "nigritude ultramarine" pages.
Are they optimzation companies that are using their own pages and running ads for the sake of it as well?

Even if I were worried, getting the ability to "opt out" of certain sites wouldn't do me any good anyway, since I would have far less of a clue than Google or Overture about which are the bad ones.
You could use your log files and click thru to the IP or search query. Found a lot of Japanese female anamie sites in the early days of content search... guess there are a lot of young women who trade currency

OTOH, I see content advertising as a branding thing anyway. The CTR is so low (like 0.2% or whatever) that it becomes very low cost advertising. When the CPM is 25 cents, you tend to overlook a lot of other factors.
I want to advertise at those prices... I am still paying $5 a Click even at reduced rates... 12-14 on the search side... but 12.5 cents a click.... where do I sign?

Much more useful would be the ability to create separate campaigns and ads for content-based advertising, so you can control bids, and target different keywords. Let me have a checkbox so the ads *don't* show on Google.com, but only on Adsense. I'd put a bigger budget on that than search advertising.
Interesting... hey anything to get the option to choose.


It is not like Google does not realize people would like this option.

If you've got a campaign at $5 a click where you've opened it up to Adsense, I think you can expect the worst, because you're in a category where a lot of people will decide that it's worth the risk to rip you off. That's where you see 'em building pages to target your keywords, and people trying to disguise the ads as site navigation.

The "big ticket" keywords in my category are also very expensive. Stuff like "search engine optimization." There are always lower-volume terms, with lower bids. The kind of stuff that doesn't bring in much traffic on searches, is very good for content advertising because you get a lot of exposure very cheaply.

Example: CTR of 0.2%, with a 10 cent bid. That's one click for every 500 impressions, at it will cost you a dime. CPM = 20 cents. That's where we're playing in the Adsense game. Even with a 50 cent bid, you're looking at a CPM of a dollar, and very low risk that your ads are showing up in the wrong place.

Not everyone is interested in branding, though. Having our company's name on a bunch of pages that mention Wordtracker makes a lot of sense, just to build awareness.

If you've got a campaign at $5 a click where you've opened it up to Adsense, I think you can expect the worst, because you're in a category where a lot of people will decide that it's worth the risk to rip you off. That's where you see 'em building pages to target your keywords, and people trying to disguise the ads as site navigation.

The "big ticket" keywords in my category are also very expensive. Stuff like "search engine optimization." There are always lower-volume terms, with lower bids. The kind of stuff that doesn't bring in much traffic on searches, is very good for content advertising because you get a lot of exposure very cheaply.

Example: CTR of 0.2%, with a 10 cent bid. That's one click for every 500 impressions, at it will cost you a dime. CPM = 20 cents. That's where we're playing in the Adsense game. Even with a 50 cent bid, you're looking at a CPM of a dollar, and very low risk that your ads are showing up in the wrong place.

Not everyone is interested in branding, though. Having our company's name on a bunch of pages that mention Wordtracker makes a lot of sense, just to build awareness.

I would be interested at those rates and may have to get the time to develop it out that way.

Hi, I just wanted to add that another measure to consider (at least for Adwords, I don't use Overture yet) could be geo targetting. You can only include countries for which you want traffic.

It's a know fact that some people in India or China are "working" in the fraud click industry.

Of course, fraud click could come from everywhere, and this idea won't be useful if you need your ads seen in India and China.

Regards

Hi, I just wanted to add that another measure to consider (at least for Adwords, I don't use Overture yet) could be geo targetting. You can only include countries for which you want traffic.

It's a know fact that some people in India or China are "working" in the fraud click industry.

Alexis,

Most advertisers are already limiting their campaigns to specific countries, but the PPCs have no way of detecting where the clicks are coming from if they use a proxy server. Since the folks who do this are already using multiple proxy servers to avoid detection, geo-targeting isn't going to be much of a defense.

On another note there are publishers who use AdSense to make a few bucks to put into other areas of marketing while restricting their direct competitors from appearing on their sites.
It seems frustrating that a publisher can restrict who appears while an advertiser cannot do the reverse.

Ouch, I didn't know about the proxy issues to avoid geo targetting, that's bad to read.

Well, then I think the only ones who can, and should, look for solutions for click fraud, are the big PPC networks.

News item:

http://www.newsfactor.com/story.xhtml?story_title=Google-Sues-Over-Click-Fraud&story_id=28609

Seems directly related to the opening post of this thread.

many people coming here often have heard me scream about click-fraud a lot...

what scares me is google being public now. as a public company they must continue to show growth, a profit and/or both.

too much publicity over click-fraud is bad for business for google. their first approach appeared to be "aloof". i have been asking my local reps about it for over 2 years and their responses are always one of the below:

- "what is click-fraud?"

- "we don't charge you for all clicks and this protects you from some of the fraud"

- "we have mechanisms in place to prevent click-fraud ...but, we can't share them with you"

- "i will talk to the engineers and get back to you..."

all of which are pretty lame. for starters, if you are in "the business" and don't know what click-fraud is, it's time to go paint fences instead. Not charging us for all clicks as a type of "shooting in the dark" methodology is just dumb and inconsistent. ...and having mechanisms in place to prevent click-fraud is a pure "snow job" as it's impossible to detect click fraud if the "frauder" is attempting to mask identity. ...and, of course, when they say they are talking to engineers and will get back you, this simply means you have treaded into discussion in which they are clueless and it's a great way to keep you at bay for a week or two while they collect their commission checks.

in a recent article on Cnet, google has dropped their own press release on the fact that they are going after a click-fraud case. i assume that they are now moving past "aloof" and into "spearhead" mode.

perhaps they think we will all think "geee, google cares. they are gonna sue the pants off someone and scare all click-frauders away"

nope.

it will grow. ...continue to grow. ...and i am resigned to believe that the only weapons we have are our own minds, industry experience and a willingness to manipulate campaigns in a manner to minimize the potential of fraud.

unfortunately, as one poster mentioned, PPC is a necessary evil for today's online businesses.

Sebastian, if you don't mind me asking: what's your point?

PPC is a necessary evil for today's online businesses.
Is advertising a "necessary evil" or simply what good businesses do to reach customers? PPC is, in my experience, one of the best ways to reach a targeted audience online. If you can't afford it, then don't do it.

Click fraud is really a non-issue for the campaigns we run. There may be a lot of fraudulent clicks in the mix, I don't know, but what I do know is that our advertising is profitable. Profitable, because we continually work on our site, our follow-up, our processes so that we can make the most out of every visitor we get.

I suspect that most of the folks worrying about click fraud have a lot better ways to spend their time, like making their website an effective selling tool. If someone can't afford to spend ten cents to get a visitor, maybe their site just isn't worth a dime yet.

...manipulate campaigns in a manner to minimize the potential of fraud.Shops have existed since time immemorial. AS has the appropriately named shoplifting. Does that mean one shoudl close shops, have people body searched on teh way in rubber glove style and the threat of theft taken so seriously that every posible precaution is taken... Or is it just a cost of doing business, and it is a far better use of time to get more customers, build better shops and stock better stuff??

Click fraud is a cost of doing PPC, and a very minor one at that in the vast majority of cases. Stuiff minimising click fraud, lets maximise profits. the two are not one and the same, and in fact minimising fraud at all may reduce profits, as the time / expense my be larger than the returns.

Move forwards, not backwards.

Sebastian, if you don't mind me asking: what's your point?

point is, click-fraud or not, ppc is still a great way to garner additional online business; but it's definitly the responsibility of the advertiser to monitor IPs and other web log information for possible fraudulant clicks. i know, i know - we are always monitoring logs anyway - but in this case, we can't rely on the engines to protect us.

If someone can't afford to spend ten cents to get a visitor, maybe their site just isn't worth a dime yet.

this is true, but .10 keywords and phrases are long gone. yes, you can create some rather obscure phrases that might carry that tiny of a CPC, but for any industry with any competition, forget it.

Click fraud is really a non-issue for the campaigns we run.

then you must sell something that is not very popular ...but for industries where competition is fierce, click-fraud is happening with brute strength. mortgages, car dealers, credit cards etc... do not even flinch at $40 CPCs. That's huge. In my industry, tickets, we haven't spent to those levels yet, but it's growing every month.

we can argue (sorry, "discuss") all day about it, but the proof is in the pudding and the writings and analysis are happening now - even in this forum - and 1 year from now, or 5 years from now we can look back on this and judge if click-fraud actually became an issue or not.

i was there as an aol teenager writing VB scripts when people said "wares" (sharing cracked software via chat rooms) was not an issue, ...now it's a multi-million dollar problem and has steered a lot of people away from writing software.

i was there when there was no spam. but what happened? enterprising individuals realised that it's a law of numbers. send enough emails and you are bound to find a certain number of dumb-dumbs who will fall for it. ...now, it's a major, if not thee major, problem on the internet.

and now, with ppc fraud, there is good money to be made in an industry where it takes real knowledge and talent to be "found out". it's the "new spam" and like email spam will take a long time to prevent. with less fear of being caught, coupled with good money to be made PPC fraud is bound to increase.

it's the perfect recipe for the internet "bad dudes". a year from now my posts will still be here and my points will become proven.

point is, click-fraud or not, ppc is still a great way to garner additional online business; but it's definitly the responsibility of the advertiser to monitor IPs and other web log information for possible fraudulant clicks. i know, i know - we are always monitoring logs anyway - but in this case, we can't rely on the engines to protect us.



this is true, but .10 keywords and phrases are long gone. yes, you can create some rather obscure phrases that might carry that tiny of a CPC, but for any industry with any competition, forget it.



then you must sell something that is not very popular ...but for industries where competition is fierce, click-fraud is happening with brute strength. mortgages, car dealers, credit cards etc... do not even flinch at $40 CPCs. That's huge. In my industry, tickets, we haven't spent to those levels yet, but it's growing every month.

we can argue (sorry, "discuss") all day about it, but the proof is in the pudding and the writings and analysis are happening now - even in this forum - and 1 year from now, or 5 years from now we can look back on this and judge if click-fraud actually became an issue or not.

i was there as an aol teenager writing VB scripts when people said "wares" (sharing cracked software via chat rooms) was not an issue, ...now it's a multi-million dollar problem and has steered a lot of people away from writing software.

i was there when there was no spam. but what happened? enterprising individuals realised that it's a law of numbers. send enough emails and you are bound to find a certain number of dumb-dumbs who will fall for it. ...now, it's a major, if not thee major, problem on the internet.

and now, with ppc fraud, there is good money to be made in an industry where it takes real knowledge and talent to be "found out". it's the "new spam" and like email spam will take a long time to prevent. with less fear of being caught, coupled with good money to be made PPC fraud is bound to increase.

it's the perfect recipe for the internet "bad dudes". a year from now my posts will still be here and my points will become proven.
It is an interesting prespective - and I agree click fraud is prevalent because people have to be found out... but the companies are more vigorious filtering for it then the email/ISP people were until just recently.

I think there will always be a small number doing it - but since there is far more criminal action involved as compared to SPAM many people will stay away as a couple more people are prosecuted....

i humbly disagree.

the bulk of the click-fraud will, if not already, be located outside the US. if you think our typical law enforcement folks are "digital dumb", travel to costa rica, taiwan or mexico all of which have very good connectivity availability.

there is such a small number of law enforcement that can grasp, understand and persue internet crime. this is why you are still getting the vi@gra emails, the porn imagery and little dumb emails that say, "your information" with a .zip file attached.

plain and simple, good internet crooks are not afraid of the consequences. i am not sure what the actual odds are, but i'd sure put money on the technical savvy of the "digi-crook" over the prowess of the "cyber-cop" any day of the week.

for every press release google fires off about taking someone to court, there are thousands laughing all the way to bank...

But Sebastian, there is a significant safeguard against this overseas clicking. AdWords advertisers can pinpoint their ad delivery by country and even region. Yes, content fraud rings can perhaps get around that, but you can, after all, disable content targeting if it doesn't perform for you.

What is all this crazy talk about someone sitting in India and clicking on my links? IP addresses from India don't count against my clients' AdWords budgets. Never have. Ads are shown to the country you specify. (Or even geographic region.) This isn't 100% accurate, but what in life is 100% accurate?

This does raise a huge point - there needs to be more flexibility and control for advertisers in how they are able to bid on different sources of publisher (content) traffic. I predict Google will come out with a major upgrade to its contextual bidding platform by Q3 of 2005, and Ov. will be forced to follow suit within 6 months of that. If they don't do this I think the crankiness level of advertisers will continue to rise, justifiably.

andrew.

good points; however, IP spoofing is very simple. remember, if these companies are going to take a risk getting involved, most will not go into it blindly.

for instance, here's a trojan-type example:

a cyber-crook could spam the crap out of millions of people with a small trojan attached that can be "awaken" (..or awoke???, i dunno grammar) to send a certain query string to google and bang a click. the click would register as coming from the 'infected' computer. ...worse, it could then remove itself or stay "alive" to simply accept another "query string" at a later time.

i think it comes down to each marketer's (for lack of better term) technical knowledge. many marketers do not have a clue what can be accomplished technically and some kinda know because they read a lot.

but for the group o'geeks, like me, deep in trenches of tech and marketing - the knowledge exists of what can be done.

agree or not - i guess we'll all see in a year or two if was right, or simply just a paranoid twit.

andrew.

good points; however, IP spoofing is very simple. remember, if these companies are going to take a risk getting involved, most will not go into it blindly.

for instance, here's a trojan-type example:

a cyber-crook could spam the crap out of millions of people with a small trojan attached that can be "awaken" (..or awoke???, i dunno grammar) to send a certain query string to google and bang a click. the click would register as coming from the 'infected' computer. ...worse, it could then remove itself or stay "alive" to simply accept another "query string" at a later time.

i think it comes down to each marketer's (for lack of better term) technical knowledge. many marketers do not have a clue what can be accomplished technically and some kinda know because they read a lot.

but for the group o'geeks, like me, deep in trenches of tech and marketing - the knowledge exists of what can be done.

agree or not - i guess we'll all see in a year or two if was right, or simply just a paranoid twit. That is interesting but I think there is a need for the session to come from a Google page to be counted for the search and I would like to see how they can write something that loads a Google page and then makes the click...

I would like to see how they can write something that loads a Google page and then makes the click...
That would be fairly easy, but I'm not sure what the point would be. I suppose that a competitor "could" do this, but is it worth the risk, and the cost? I never click a competitor's ad on Google, because that drives *my* click cost up too, by raising their CTR and pushing their ads upward.

The "low hanging fruit," when it comes to click fraud, is contextual advertising and Adsense. If people want to steal from Google's advertisers, that's where they'll do it. The simple way to avoid it is to avoid contextual advertising, if you're that afraid of it, and let your competitors own that channel.

It still comes down to whether or not your advertising is profitable. No matter what the CPC is for your ads, you're either making money or you aren't. If the CPC is $40, and someone else is making money, they'll keep advertising. If you aren't making money at that price, then you can either stop bidding at that level and get the traffic you can afford, or you can work on things that drive higher conversions and profitability.

That is interesting but I think there is a need for the session to come from a Google page to be counted for the search and I would like to see how they can write something that loads a Google page and then makes the click...

Aussie ...your a smart dude and i respect your opinions, but your comment above shows a limited understanding of IP and script writing. my example would be ultra-simple and worse, could even be written to "write itself" in different variants depending on attributes of the user's machine.

:-) ...no joke.

what can and can't be done really comes down to the technical knowledge of the person involved in the discussion and quite honestly, goes just a tad bit off-topic.

i will bow out of the discussion with this last comment:

in whichever and whatever product, click-fraud exists and is growing. click-fraud will probably continue to grow. as long as there is someone to benefit from a click and thus, someone to lose from a click - it will exist.

some marketers and site owners need not worry, it may never affect them ...and some in some crazy competitive industries do need to worry.

technically, it's very easy and simple to mask identity.

i guess each person, based on individual belief, will act accordingly.

i am just glad to have a "forum" in which to discuss it because i believe the topic is here to stay....

I have to admit, it concerns myself a little as an AdSense publisher. After all, some of us admin multiple sites - and if third-party click-fraud were affecting our own clickthroughs - even on just one site - then what is to stop Google scapegoating the publishers as the source of the fraud, and wiping out our entire AdSense accounts?

I *presume* that AdWords is Google's major source of income, so they'll put major resources into protecting the integrity of the system so it continues to be well-used.

However, I can't help but agree with the point Sebastian raised - that this is a problem now, and could become much worse tomorrow.

andrew.

good points; however, IP spoofing is very simple. remember, if these companies are going to take a risk getting involved, most will not go into it blindly.

for instance, here's a trojan-type example:

a cyber-crook could spam the crap out of millions of people with a small trojan attached that can be "awaken" (..or awoke???, i dunno grammar) to send a certain query string to google and bang a click. the click would register as coming from the 'infected' computer. ...worse, it could then remove itself or stay "alive" to simply accept another "query string" at a later time.

i think it comes down to each marketer's (for lack of better term) technical knowledge. many marketers do not have a clue what can be accomplished technically and some kinda know because they read a lot.

but for the group o'geeks, like me, deep in trenches of tech and marketing - the knowledge exists of what can be done.

agree or not - i guess we'll all see in a year or two if was right, or simply just a paranoid twit.
If this is just going to be about geek chauvinism, then I decline to play. Being deep in "the trenches" is neither here nor there, it's just another place to be. Calling marketers dumb is Slashdot material. Is that what this forum is turning into, on an AdWords thread no less?

You're making the grave assumption that you can vacuum the moral and legal implications of these actions out of the discussion. News item about a mile from my house: a businessman (who knows why) leaves the bank with $50,000 in cash, and is held up. I've read enough to know that the reason the thieves got his money is because highly sophisticated technologies called "handguns" could splatter his brains all over the pavement. Because you can do it doesn't make it right. You might as well have me cowering under my bed because anthrax might come to my door, or the terrorists will fly into the CN Tower this time.

This debate is NOT about marketers' lack of sophistication, is it? Is it a "geek" thing to show "what I can do" by committing a felony with one's knowledge of IP spoofing? Or is that a "criminal" thing? I realize you are trying to do the responsible thing by explaining to us what is possible, but what about the other side, where a company of good will like Google has very little interest in seeing its advertisers pay for bogus clicks? True, you can't catch foreign crooks, but you can issue refunds. I have heard a lot of hearsay about Goog. and Ov. being difficult to deal with in this area; perhaps, but I manage as many accounts as anyone, and just haven't seen the rampant problems you warn of.

So again, I'm not quite sure what the point of all this is other than to say "marketers don't know....x" or "the sky is falling." We all acknowledge that there are flaws in the pricing of online advertising. I wish the Audit Bureau of Circulations or someone like that would get more involved in PPC. But they'd need to be paid to do so, and that would take a full-industry commitment.

I have to admit, it concerns myself a little as an AdSense publisher. After all, some of us admin multiple sites - and if third-party click-fraud were affecting our own clickthroughs - even on just one site - then what is to stop Google scapegoating the publishers as the source of the fraud, and wiping out our entire AdSense accounts?

I *presume* that AdWords is Google's major source of income, so they'll put major resources into protecting the integrity of the system so it continues to be well-used.

However, I can't help but agree with the point Sebastian raised - that this is a problem now, and could become much worse tomorrow.
Come on, now, don't assume. It should take about three seconds to look this up. Everyone on this forum needs to know approximately how much of Google's revenue comes from advertising, just so we understand the company we're always talking so much about.

Answer: it's in the neighbourhood of 95%.

I’m a big fan of PPC but the fraud needs to be dealt with. Clients often ask me about click fraud when proposing PPC, i hate having to tell them that PPC fraud is so easy to commit.

I recently had some click fraud on one campaign and was refunded by Google without asking. The money was a welcome surprise but shows i need better tools to help detect this. Anyone know of a log analyser that can highlight/detect click fraud?

How many of you have detected click fraud on your campaigns and were you able to convince google?

As far as committing click fraud goes people don’t need sophisticated tools, IP spoofing or programming skills. It’s as simple as downloading one of the many free proxy tools and a big list of anonymous proxies. Although your professional fraudsters will have their own tools for this. I know Indian labour is cheap but why bother paying people to click on ads, a bot could do the job for free 24/7.

One of the tracking programs I use has a click monitor also.... it has helped keep an eye on fraud.... KeywordMax.... and it offers a free full functional trial for a quick sweep.

I recently had some click fraud on one campaign and was refunded by Google without asking. The money was a welcome surprise but shows i need better tools to help detect this. Anyone know of a log analyser that can highlight/detect click fraud?

How many of you have detected click fraud on your campaigns and were you able to convince google?

Paul,

The tools that the PPCs have to detect fraud are far more sophisticated than anything we can do with log file analysis. Your logs won't tell you that a specific Adwords account is getting an unnatural number of clicks from:
1) The same ISP's dial-up accounts
2) The same group of 4000 proxy IPs

I've seen more credits from the PPCs doing it on their own, than from the couple times I've raised an issue. There are adjustments made every month on all the contextual ad campaigns I monitor,but I've never had any problem with Overture or Google when we can document a problem. With the so-called "second tier" PPCs we've had enough problems that we don't do business there any more.
As far as committing click fraud goes people don’t need sophisticated tools, IP spoofing or programming skills. It’s as simple as downloading one of the many free proxy tools and a big list of anonymous proxies.
The problem with click fraud is that the criminals (wire fraud is a felony) doing it are trying to make money. Which means that they target Adsense (in Google's case), or open affiliate programs (in the case of other PPCs), so they can get paid.

It's probably not as simple as grabbing a public list of open proxies, because the PPCs can get those too. With that said, I'm sure it's possible to fool the PPCs with the right scheme, at least for some period of time. But there's no other form of advertising where we can control how much we spend and target it so effectively.

Dealing with click fraud beats the heck out of advertising in a publication that overstates their true circulation by an order of magnitude, buying "verified" postal mailing lists that are 50% undeliverable, and all the other ways you can get burned on advertising.

The tools that the PPCs have to detect fraud are far more sophisticated than anything we can do with log file analysis. Your logs won't tell you that a specific Adwords account is getting an unnatural number of clicks from:
1) The same ISP's dial-up accounts
2) The same group of 4000 proxy IPs


This is the kind of tool i am looking for, considering the amount of money involved i thought one would exist. A few thousand for some software is within my budget. Not keen on using the online solutions that are available as their pricing structures does not suit the many small campaigns i run.


It's probably not as simple as grabbing a public list of open proxies, because the PPCs can get those too.

Agreed, but I wrote a very simple app to test click fraud and used a proxy list freely available on the web, i then used the app on one of my dummy campaigns to see what detection methods the PPC firsm were using. All the automated clicks i generated were registered without any later refund. Can only guess that the actual volume of fraud had to be significant(more clicks, longer period) before it gets noticed, its this fraud that tries to stay under the radar the i’m interested in tracking/detecting. Just before my refund from Google i noticed the unusually high CTR, but normal log analysis tools were not much use.

Guessing many firms have developed in house tools to detect fraud and probably don’t want to reveal the methods they use – maybe have to write my own but fiding the time is a problem.

cheers

Paul,

Comparing the IPs vs. known open proxies is one way to go, but it's just very hard to find a pattern in *your* logs, when you may be one of a thousand advertisers who are being victimized by a particular criminal. I haven't found any of the services to be cost-effective vs. what we believe is actually going on within our campaigns.

The best defense against click fraud is good measurement of ROI. Given that it exists you have to tighten the numbers and know what really is converting and for how much.
Though it does not elimate it ... so long as the price is right you can avoid it taking a bigger toll than it could.

click-fraud is happening with brute strength. mortgages, car dealers, credit cards etc... That is no suprise, these are the same industries that are the most notorious for spamming etc. If they are now focusing on each-other, all the better.

I have run AdWords for at least 2 years now. At first I was forking out about $70.00 USD per day and getting about a 5% ROI. After constant tweaking, monitoring, geo tatgetting, turning off content match and not paying a high price per click I'm now getting a >100% ROI on most days.

For me, Google AdSense is by far superior to any other PPC out there. So much in fact that I now only use Google AdSense.

Like others, I'm sure I too have been a victim of click fraud. But it's very much a 'practice makes perfect' tool. In other words, you may intially loose money on PPC, but in time that can change for most.

Come on, now, don't assume. It should take about three seconds to look this up. Everyone on this forum needs to know approximately how much of Google's revenue comes from advertising, just so we understand the company we're always talking so much about.

Answer: it's in the neighbourhood of 95%.
Should every AdSense publisher publisher therefore be confident that, when they see what appears to be suggestions of clickfraud on any of their sites, that because 95% of Google's revenue is from AdWords programs, that therefore it's the clickfraudsters that will be punished rather than publishers? That was my point - it would have been nice for you to jump on that, rather than my diplomatic use of language.


This debate is NOT about marketers' lack of sophistication, is it?

Indeed it isn't - yet you managed a few paragraphs trying to drag this thread down in that manner. Your comments at Sebastian suggest that you're more interested in looking like a good marketer, rather than acknowledging that the whole click-fraud area is one of growing concern. I *presume* that is not Google's attitude. But as has been pointed out, this isn't a simple problem to address. And Sebastian warns that it's going to get worse.

The core question is as to what degree can Google and Overture, et al, combat the issue - at least to the public degree of maintaining confidence with the sponsors and publishers.

Actually, the "core question" can be seen in the Title bar just above :)

Actually, the "core question" can be seen in the Title bar just above :)
I guess the "core answer" is no, because we're not all getting ripped off.

Some of us are affected by click fraud, though, but IMO it's already factored into advertisers' bids. I presume that whatever click fraud exists in a particular market, advertisers are adjusting their bids accordingly.

That's how markets work. Advertisers will pay as much as they can afford to, given their particular situation, however they measure ad performance (absolute ROI, cost per action, etc.)

If click fraud were completely eliminated, advertisers would bid more per click for the higher quality traffic, and any short-term benefit in terms of ROI will be absorbed by higher click charges.

This isn't a defense of click fraud, but it may be a better way of explaining why I'm not terribly worked up about it.

Quite frankly, I think "dumb searchers" are a bigger problem, as evidenced by the fact that I can't run ads for terms like "keywords" or "search terms" without adding negative matches for "type," "AOL," "here," etc. If I didn't use those negative matches, I'd get referrals for "type AOL keywords or search terms hereviagra" and stuff like that, because so many searchers just click the first thing they see on a SERP without even reading it.

BTW, there's a free PPC tip there for all of you drug pushers, bid a nickel or a dime for "hereviagra" and you'll get all the traffic you can handle, dirt cheap. In fact, "heresearchterm" is a nice cheap target for Adwords across the board.

The disabled visited color on clicked links probably accounts for a good 10% of your PPC campaign, standard web navigation is beyond adwords?

If they can't accomodate standard web courtesy I doubt they care about fraud.

When I click any AdWord the link goes a visited color. It only goes back to unclicked via Refresh or another search.

George Reyes Google CFO said only a couple of days ago that something has to be done about this and it threatens their business model.

The problem is also so that industry experts like Jessie Stricchiola reckon that it could be much more than 20% of all clicks are Fraud.

Now if Jessie is right stopping click fraud is going to hurt googles pockets and deep, also remember Google's primary defense has always been to refund advertisers if they complain and only if Google can see evidence that Bad clicks have occurred ....

The real Problem is the advertisers have to prove that fraud has happen...

Advertisers should go and find good people to help them detect the click bots, it's not that hard we have install a good in house detection script of couple of clients...

also Google knows of another solution which I talked to Matt about in Vegas which he is going to look into.. but you still have all the other PPC engines to sort out as well ... oh hum

DaveN

The problem is also so that industry experts like Jessie Stricchiola reckon that it could be much more than 20% of all clicks are Fraud.
That may be true, although I suspect that number is much overstated in terms of the top two PPCs. I'd prefer to see an assessment from someone who isn't in the business of auditing PPC campaigns. It's hard to find an independent voice, though, since only the people down in the trenches (like Jessie) have a good view of the problem.
Stopping click fraud is going to hurt googles pockets
I don't believe that this is true, as stated in my earlier post regarding markets. If 20% of the available inventory is fraudulent, and it goes away, the price of the remaining 80% will go up and revenues will not be seriously affected. Remember, the demand in the market is only for the non-fraudulent clicks.

Beyond that, if the threat of click fraud vanished, how many more advertisers would begin running contextual campaigns on Adwords? Eliminating click fraud would be a great thing for the honest PPCs like Google.

Stopping click fraud is going to hurt googles pocketsPerhaps, but it maybe a case of those that are victims are already adding to the kitty unknowingly. In other words, the unknowing victims are going to foot the bill.

It will only hurt them if the incremental cost of stopping fraud (software, servers, people, etc.) exceeds the revenue they add by having a better service for advertisers.

Let's say that I am an advertiser, and I am willing to pay $1 for a quality visitor (let's define that as an average 'real' visitor from Adwords) because that hits my magic number for ROI. If 20% of the clicks are fraudulent, then for every 1000 visitors that Adwords sends me, 800 will be quality visitors. So I am willing to pay $800 to get those 1000 visitors, or 80 cents per click.

Now, let's say that Google magically eliminates fraud. Instead of 1000 visitors, Adwords now sends me 800 visitors, because they got rid of 20% of their inventory. But they're all quality visitors, so I am still willing to pay $800, which is now $1.00 per click. Google's going to get $800 out of me either way, because I'm going to get the same profit either way.

Or will I be willing to pay more, since eliminating fraud also cuts the cost of managing my PPC campaign?

Or, using arbitrary figures, if Google profit $100,000 per day from AdWords, $20,000 is the result of click fraud and only half get refunded, they have $10,000 per day to put toward technology to stop click fraud.

Or, using arbitrary figures, if Google profit $100,000 per day from AdWords, $20,000 is the result of click fraud and only half get refunded, they have $10,000 per day to put toward technology to stop click fraud.
you are only going to wipe out so much click fraud and eventually as you wipe out more prevention is going to start costing more than it saves.

http://www.bartleby.com/65/di/diminish.html

certainly the perception of a fraud free network may inspire additional advertising, but it will cost a bunch of money to stop the fraud. keep in mind that some webmasters and fraudsters are exceptionally technically astute.

you are only going to wipe out so much click fraud and eventually as you wipe out more prevention is going to start costing more than it saves. Agree that they may never completely stop click fraud, but I doubt anyone (outside Google) could state, with any certainty, that it will start costing more than it saves.

Dave,

The law of diminishing returns is called a "law" for good reason. You can't get around it. Eliminating fools who click on their own Adsense listings can be done very easily. Those using known open proxies to do the same, it's a little harder, but not too bad. That's the low hanging fruit.

From there, it gets more difficult, and more expensive, to identify each fraudulent click. They'll never get it all, and at some point it becomes too costly to take a bigger bite out of fraud. Some folks have already named a few schemes (virus/worm, thousands of dial-up accounts, etc.) that would be pretty much undetectable at the network level.

Possibly the best thing Google has done to combat fraud is implementing smart pricing, where the value of the Adsense clicks decreases if the traffic doesn't convert to sales.

The law of diminishing returns is called a "law" for good reason. You can't get around it Yes, but it's not correct to apply it as a 'blanket law' and assume it applies in all situations. I totally agree that it likely not possible to eliminate all click fraud, but I don't/wont assume that Google will hit the point of diminishing returns before they have done all they can, or need to.

but I don't/wont assume that Google will hit the point of diminishing returns before they have done all they can, or need to.
what other logical spot would you suspect Google to want to stop at then?

The happy medium. I doubt they will keep going until they hit the point of diminishing returns. It's not exactly a target to aim for is it?

It's not exactly a target to aim for is it?
for most businesses it is.

We will have to agree to dissagree on that one.

The Law of Diminishing Returns (anticipated by Anne Robert Jacques Turgot and implied by Thomas Malthus in 1798) states that increasing one variable of an equation while keeping the rest of the variables constant will eventually yield a result opposite the intended purpose of the variable change. In plain English, it simply means sometimes, you can push an idea too far. Source: http://www.clickz.com/experts/crm/crm_strat/article.php/3365481 Bolding added by me. That really doesn't sound a target a business would aim for. That is, a business would normally aim for the most profitable point.

That is, a business would normally aim for the most profitable point.
which is right before they get to the point of diminishing returns. :) hence the definition, right? in fact the the law of diminishing returns states that there is a most profitable point.

we are really off topic, back on topic :)

Google AdWords is a great program and we are not all getting ripped off. to some extent for many of us click fraud is factored into our bid or click price

which is right before they get to the point of diminishing returns. No, not necessarily at all. One simply cannot use a blanket statement like that as say it applies to all business in all cases.

in fact the the law of diminishing returns states that there is a most profitable point. Of course there is a "most profitable point" but that is not the same as saying (as you have) that the point of diminishing returns (or just before) is the "most profitable point".

In fact, the point just before, or at, the point of diminishing returns is more likely the 'break even point'. Fine for charities, non-profit organanizations etc but not a smart target for a for-profit business to aim for.

Are there problems... obviously...
Will solutions eventually be found.... obviously....
When they are will the prices for the terms go up... obviously...

Are there problems... obviously...
Will solutions eventually be found.... obviously....
When they are will the prices for the terms go up... obviously...

well said.

but I will add that in the meantime there are a LOT of people getting ripped off in the PPC world.

Yes they are and "obviously" is just a word which may, or may not, end up being true.

In the past few days I've heard some pretty incredible horror stories from people paying the big guys a lot of $$. Blatant fraud and the only response they get is "we can't prove it".

refunds are like pulling teeth, and nobody wants to open the floodgates.

but I will add that in the meantime there are a LOT of people getting ripped off in the PPC world.
But who by? Their agency that buys crap ads? The Search engine that allows click fraud? There web developers that charge $250 an hour to build PPC landing pages or the idiots that look for "google" and click on a PPC ad for "Google AdWords Management"? It would help if you qualified your statement a bit better, so that I can beter understand what and where you believe the "LOT of people" are being so badly hurt> It may even be a business opportunity...

FWIW, IMHO there is likely to be more money wasted every day on poorly put together PPC campaigns than click fruad costs in a year. Rather than putting the horse before the cart, create a good quality PPC campaign FIRST, and then tweak for less costly factors later.

How many people are using click fraud programs???
These will give you the logs that you can use to argue for refunds from any search engine... it gives them something to look up as well.
Obviously there are dynamic proxy servers etc that make this a little conflicted but they too can be investigated against publications that are referring the clicks and a bunch of ways.
Look at those to start a counter measure.

It would help if you qualified your statement a bit better, so that I can beter understand what and where you believe the "LOT of people" are being so badly hurt> It may even be a business opportunity...

A lot of people are getting ripped off by competitors, ex employees and the like clicking their ads, and the PPC companies are not terribly helpful about giving refunds even with blatant examples of fraud being shown to them. If everyone had a close look at their logs, a large percentage of them would see patterns of clicks from the same people day after day week after week - and they are getting charged for almost every single click. If you are in a business paying $1, $5 or even $20 per click, this can add up very quickly.

IMO, you should not be charged for clicks that come from the same exact person on the same exact computer every single day, but almost every single time you will be.

Forget click abuse, google has created click abuse with their last change in the way keywords are handled.

What ever happen to the good old 'i pay more than you so i get listed on top' days?

Now I need to practically click my own ads so google uses the ctr to deem my ads RELEVANT so it will actually SHOW my ads. They are deminishing the quality of their paid ad network to increase revenue on their end.

My god, what good is paid search if they are using CTR to determine relevancy. It's obvious the ones near the click zone for that user will get clicked more as long as the ad is targeted.

My campaign was fine until googles last update of this new TRIAL and ON HOLD stuff. Now I cant even get impressions in very obvious, clear, targeted 2 or 3 word phrase ads! No more than 5 impressions and my keywords phrase is on trial or hold. What gives?!?!

I've officially paused my googe adwords campaign.

Its not a waste if you analyze your adwords and set a lower budget for testing. Don't think you can become a millionaire from Adwords. Not everyone is successul.
Manny~