Eric Schmidt’s recently discovered comments (via Donna Bogatin) about click fraud have got a number of folks headscratching. I’ve long concluded that for advertisers, click fraud is a tax, one that they will bear until it gets too onerous (in other words, until the cost of fraud outpaces the return of CPC advertising). Journalists lick their chops at perhaps someday breaking this story wide open. But if there is a larger story here, it will be broken by a deep throat inside Google, not by anyone else. That’s where the truth can be found.
Fact is, we have no idea how much click fraud there really is. Recent reports put it at 14%, but whatever it is, it’s a real issue in the minds of Google’s customers, and making light of it sounds entirely off key. By calling the problem “self correcting” Google forgets that it profits from fraud, and that’s a PR problem, if nothing else.
From Donna’s post:
Schmidt indicates, however, that Google engineers think it is “great fun” to try and get ahead of click fraud:
“But because it is a bad thing, because we don’t like it, because it does, at least for the short-term, create some problems before the advertiser sees it, we go ahead and try to detect it and eliminate it.”
It sounds like Google is “doing the advertiser a favor,” rather than attacking and acknowledging a real issue in the minds of its customers. Seems to me that beating fraud isn’t about having fun, it’s about insuring your customers’ trust in your business.
Eliminating click fraud would be to the long-term benefit of the search engines as well as (obviously) to the benefit of their advertisers. Not just in PR terms, but also because companies would be willing to pay more for fraud-free clicks.
If, say, 14% of clicks are fraudulent, customers should be willing to pay at least 15%, and more likely up to 20% more for clicks without fraud. The reason is that click fraud is an uncertainty. It may be 14% on average but for any given ad, it could be 5-50%.
With the same ROI, people will pay less when it’s an investment under uncertainty. That’s a key reason companies are willing to pay more for PPC ads: that they have much less uncertainty than CPM ads.
The more you can reduce the uncertainty, the more you can get for the ads (or any other investment you are trying to get people to pony up for). That’s also why pay-per-action (e.g., affiliate programs) should lead to the highest fees in the long term.
“Schmidt indicates, however, that Google engineers think it is “great fun” to try and get ahead of click fraud:”
Some of those engineers will undoubtedly be invited on the “party plane”. Oh what fun to be a Googler these days.
I’m probably just jealous, I really haven’t figured it out. I was listening to KGO radio from S.F. over the weekend and the head Google chef was on a program and he mentioned to the host that there are “dozens” of cafe’s in the G system now. He was mentioning his new signiture “cafe” named “7”, or “Seven”, I don’t know which, doesn’t matter.
If history is any guide these excesses will come home to roost one day. I hope not though, there should be at least one workplace somewhere where fantasy is lived out. A place where party planes sit on the tarmac, where happy engineers frolic and where the food is served to order 24/7.
One Factor that was NOT mentioned in any of the quotes…..was the word “PRINCIPLE”…..
Just like Google Engineers are extremely focussed on eliminating SPAM or Black Hat SEO from the SERPs, to the point of Banning entire sites, should not the same principle apply here in PPCs??????
Regardless of the ROI of PPCs, companies like Google Yahoo and MSN have an Ethical obligation to use their technology to address this flaw.
This is fundamentally no different from Robbery which in the USA is a felony, or various White collar embezzlement, Enron-type crimes.
Google should NOT be surprised by any future Class-action lawsuits, where Schmidt’s quotes will be used by those Plaintiffs.
Calm down SE Web, and let free market economics save us all. Jakob’s right – Google’s been *on top* of click fraud relative to everyone else in the industry, and they’ve been on top of it because they know it’s costing them money in the form of lower willingness to spend.
We manage more PPC than anyone on the planet, and our data suggests that there is not a single top 50 source of traffic on the internet that converts better than Google traffic. Until that changes, Google’s OK by me and hundreds of thousands of other advertisers.
Since we’re looking at Donna Bogatin’s latest post, I just thought I’d jump in here and highlight the post Donna wrote earlier this month following Yahoo!’s settlement of the class action click fraud litigation with Checkmate Systems: http://blogs.zdnet.com/micro-markets/?p=183
Click fraud is an issue we take very seriously, and it’s something we’ve recognized as a challenge since we started out as Goto.com in 1998. We’ve built up a very robust clickthrough protection system since then, and as you’ll read in Donna’s post, we’ve committed to some significant initiatives to help share our learnings and provide advertisers with more transparency about click fraud.
Click fraud is a losing battle, and it’s good to know that even billionaires are susceptible to extended Freudian slips.
Protections are great and it’s always good to see customer focus, but are IMHO life support until we reign in the new master plan. (maybe I got that idea from reading The Search, anyways…) I hope Eric slips his thoughts there some time soon, inquiring minds want to know.
While I agree with Chris to an extent, I’m just going to point out some logic flaws in Eric Schmidt’s reasoning:
He assumes that all advertisers receive the same amount of click fraud across a keyword equally, and the free market will adjust the prices to include it. However, if 1 or 2 advertisers are “targeted” by click fraudsters, if they adjust their prices down they will lose position, even if they were more relevant to the user.
Vinny has a point.
SE web has a point.
But this fraud, lets call it what it is, this has to be dealt with just like any other fraud, need to get law enforcement involved. Let them clear out the thief’s. This should give the industry a more warm and fuzzes with advertisers
When people talk about click fraud I think it’s worth separating out the channel element. If you’re talking basic AdWords click fraud where a user or users clicks repeatedly or automates a clicking process, this has got to be more malicious than anything as the individual or robot that clicks has no financial incentive to do so – spite is the only motive I can think of.
However, if you’re talking AdSense then there is indeed the possibility of a large element of fraud as the host of the web site benefits financially from the ads and thus has a vested interest in seeing as many clicks as possible.
I run paid search for a division of a Fortune 500 company and we bid low on AdSense on purpose, about 1/5th of what we bid on AdWords for the same keyword.
We see higher, much higher, CTRs on Google and AOL-hosted ads compared to AdSense. That tells you something.
Click fruad earns Google money, and thwarting it loses them money, since they have to pay back the advertiser. Does Google have the proper motivation to stamp out click fruad?
Controlling click fraud is about filtering out as much fraud as possible in real time to reduce the economic incentive to perpetrate fraud and identifying and eliminating the bad behavers. Doing both of those protects the advertiser and rewards good publishers for good traffic. Google, in my opinion, does a pretty good job of it, but they could do better.
Schmidt’s view of the corrections of an efficient marketplace are right in principle, but Google isn’t a free market, so his statements don’t apply to the Google specifically.
Google is in the position to artificially sustain an “acceptable” level of fraud because they own more information than do the other participants in the Google economy. I’m not saying they are doing so nor that it is in their long term interest to do so, just that it is their current luxury. Furthermore, they may be unaware of fraud that’s operating outside of the lens of their algorithms and inside the cloud of conversion noise. I have personally seen traffic that we’ve filtered that they are apparently accepting and paying out for.
Technological innovations are also a market response to a given phenomenon. Had we not noticed the prevalence of click fraud in PPC a few years ago, we wouldn’t have developed our technology to combat it and protect our network.
It’s our position that value sustains the long term growth of the industry. It’s in our best interest to act on suspicious clicks, even if doing so means giving away a free click to an advertiser, because doing that is what ultimately drives higher conversions and more robust bid environment for our publishers.
Click fraud is much more offputting than just “uncertainty” for advertisers. Avoiding click fraud translates into a much more basic human emotion – that of avoiding being screwed.
If I discovered that someone had copied my house key and that my house was being used as a brothel while I was at work, even if it had been going on for years and they’d not damaged anything, I’d still change my locks and call the police. If it happened again, I’d move house, despite the cost.
Issues such as this highlight the tremendous conflicts of interest between Google customers and Google shareholders that currently exist at the heart of how Google currently operates. This isn’t just about being half-hearted about tackling click-fraud but also about numerous other conflicts at the heart of adwords (e.g. differential pricing based your current spending levels).
Ultimately it might be that Google recognises that it is truly in its long term interest to allow 3rd parties to provide the advertisers’ interface – thus removing many of these conflicts at a stroke. (The current API is incredibly restrictive).
Whatever Google decides, they’ve definitely got the culture to resolve these conflicts – the challenge is to address them while the public’s trust remains high.
Just for the record – we find that AdSense (Contextual Ads) have a high levels of Click Fraud, but Google Adwords typically does NOT have click fraud levels above 10%, which is debatable in terms being statistically fraudulent. I do agree with Eric Schmidt on the “suck it up” point in so far as Search Targeted ads are concerned – most of the “fraud” is possibly related to the inability of most merchants to accurately track click, sales & returns, and tie them back to a user.
In calculating the 14%, that is averaging the high fraud levels of AdSense ads with the low fraud levels of the Search Network ads. It’s misleading. If people actually monitored their AdSense traffic for conversions, they could easily pick up the fraudulent sites, and exclude them from the ads.
I think the whole story is overblown on the Search Ads, but the Adsense ads needs to be looked at seriously – and the two should not be combined.
No more worrying about clisk-fraud since I use very smart script. But still I’m not sure if one day I receive well know terrible message from Google… One thing is sure – I will not be invited on the “party plane” 🙂
Me as a adword advertiser is more concerned with the variable keyword pricing in the name of “quality score” & inconsistency than clickfraud.
Me as a adword advertiser is more concerned with the variable keyword pricing in the name of “quality score” & inconsistency than clickfraud.
I guess I’ve never understood the “tax” analogy to click fraud. When taxes are paid, they generally (but not always) go to the greater good of the population paying them. Taxes pay for shared infrastructure (roads, highways), public services (schools), and the like. There is also (ideally) accountability for the taxes, ie. it is possible to review the expenditures, project status, etc.
None of these are true with click fraud. It’s criminal activity.
I don’t qutie understand why the click fraud numbers are singled out for accuracy check. As far as I can tell, most of the stats on the web are murky, including number of page views, unique visitors, click through rates, etc. Has any measuring/reporting entity opened the method it uses to collect data and estimate the metrics? Has anyone seen a credible statistical analysis of the bias in those estimates?
Great furtunes are made/lost based on these numbers. M&A offers of huge amounts are placed based on those numbers (facebook, bebo). Yet, no one really knows how reliable those numbers are. So the click fraud numbers are in good compnay. It’s about time that the wisdom of the crowd is directed to demand transparncy.
Michael, you are correct that reported numbers of pageviews, CTR, etc. are not always reliable. My guess is that click fraud generates more attention because more people consider themselves victims of lost revenue due to it.
“Google engineers think it is “great fun” to try and get ahead of click fraud”.
It looks really they have “great fun”. I can’t believe they cannot get ahead of click fraud. For me it is rather an open door to eliminate those publishers who do not supply potential y future clients.
The advertiser will not pay any more or less if click fraud is reduced, or entirely eliminated. The price is set on ROI and the market determines the price. In other words, without click fraud, the ROI will increase, thus increasing the competition for the keyword or phrase.
Eliminating click fraud is clearly in Google, MSN, or Yahoo’s interest. It’s easy to get lost if you look at just the HF keywords, when you look at the mid market CPC customers you’ll see low market penetration, and a higher ROI will drive that market.
Always when I hear about click fraud I’m getting ..crazy.
I can’t believe they cannot get ahead of click fraud.
This is fundamentally no different from Robbery which in the USA is a felony, or various White collar embezzlement, Enron-type crimes.
Whatever Google decides, they’ve definitely got the culture to resolve these conflicts – the challenge is to address them while the public’s trust remains high
I have lost a lot of money on chetaing AdSense users :/