[adwords.blogspot.com...]
Today, we’re announcing the launch of a new AdWords feature enabling advertisers to have a much more detailed picture of invalid click activity in their account. The metrics of “invalid clicks” and “invalid clicks rate” will show virtually all the invalid clicks affecting an account.These clicks are filtered in real-time by our systems before advertisers are charged for them. The resulting data will of course differ from one advertiser to the next. In addition, a much smaller number of invalid clicks may also be credited to advertisers’ accounts after-the-fact, as the result of a publisher being terminated from the AdSense program for invalid click activity. These will appear as account-level credits.
[edited by: engine at 8:24 am (utc) on July 26, 2006]
[edit reason]
[1][edit reason] added quote [/edit] [/edit][/1]
It's hard to see how third-party verification would change the minds of advertisers or critics who have preconceived notions about a PPC vendor's approach to click fraud.
Really, I think it would help A LOT.
What's adequate may be in the eye of the beholder, but one unbiased, court-approved "3rd party" has already concluded that Google's "efforts to combat click fraud are reasonable."
lol, that "expert" was approved by both parties to the click fraud suit in oder to settle the case. The case will settle but one thing remains:
THE FUNDAMENTAL PROBLEM
"Therefore, it is simply impossible to identify true clicking intent for certain types of clicking activities and, therefore, classify these clicks as valid or invalid."
Savvy advertisers know that the google report is just a best guess and considering that google let "double clicks" slide for years (doh!) the "expert" report is more of an indictment than an absolution, nothing presented engenders confidence in G$$Gs willingness or ability to fight fraud.
Also, some people are going to be unhappy no matter what Google does. Why should Google even waste time in trying to satisfy a tiny minority of unhappy, vocal, and often vituperative AdWords advertisers and AdSense publishers? (Anyone who's ever worked in a service business knows that some clients just aren't worth the trouble they cause.)
Again, the bottom line is return on investment.
Yes, the perfect economic solution, let click fraud happen. Its pretty clear that is the solution G$$G wants to put forth. We get rich while your ROI goes in the toilet!
Unless I'm missing something, this suggests to me that one person and sit there and click on a keyword over and over and I get billed each time. There is no way of knowing if this is legitimate traffic or a competitor doing this.
BTW don't criticise google too much till you see how little yahoo does. I am waiting a month for a reply on an obvious fraud on yahoo.
I think this is a great little feature to settle clients fears from reading one too many Wall Street Journal articles, but in reality is more or less a gimmick.
I applaud Google on this move and I think it is a step in the right direction, but as people have noted in this thread, the only real answer is to come up with a set definition for what an invalid click is and get a third party vendor that can get in there and get you the information to receive a timely refund. We need more powerful analytics packages and increased cooperation from the search engines.
At the same time its more than you can say from Yahoo or MSN Adcenter.
I still can barely track what valid clicks I'm actually paying for with Adcenter
[edited by: arrowdevelopment at 7:12 pm (utc) on July 26, 2006]
In the end, advertisers will know if Adwords provides a better ROI than Yahoo, or radio, or bumper stickers, or direct mail, or whatever. Publishers will know if Adsense provides a better payout than an internal salesforce, or Kanoodle, or a full-time job.
Google's long-term business depends on the relative effectiveness of efforts occurring inside the Black Box. Because that's where the balance between the interests of Advertisers, Publishers, and Google ultimately occurs.
Everything will stabilize. Brin will buy a new airplane. etc.
TypicalSurfer, thanks for proving my point. :-)
no problem, to illustrate how your point (let click fraud happen) on the "economic solution" works, I'll walk you through the latest "craze" in competitive adwords management.
Goal: increase ROI in an ever increasing fraudualent market.
How: Run Adsense sites in your KW area, run fake traffic at your adsense sites, click on your competitor ads, fund your adwords budget with bogus revenue directly tapped from your competitors, voila! Your ROI is now acceptable. The perfect economic solution at work.
Sorry EFV, your solution is sadly lacking.
I have 57 accounts with an average of 10.10% invalid clicks. My campaigns with the highest CPC have the highest percentage of invalid clicks. Is this because of AdSense publishers cashing in on high priced clicks or is it because my competitors are running up by costs?
Is this because of AdSense publishers cashing in on high priced clicks or is it because my competitors are running up by costs?
It could be both, in the example I cited competitors are now working adsense to not only deplete competitor budgets but to fund their own adwords campaigns. There is no black and white "either or" type of fraud, there are many levels of grey. If you think G$$G can catch fraud by simply observing/analyzing click streams I've got a bridge you might be interested in buying.
no problem, to illustrate how your point (let click fraud happen) on the "economic solution" works, I'll walk you through the latest "craze" in competitive adwords management.
But I never said "let click fraud happen."
I did say: "Why should Google even waste time in trying to satisfy a tiny minority of unhappy, vocal, and often vituperative AdWords advertisers and AdSense publishers?" Perhaps I should have inserted the word "dishonest" after "vituperative."
If Google really wanted to provide advertisers with a tool to fight click fraud, they would allow the advertisers to view the IP addresses of the clicks. If I could block the IP address of my competitors, I would have a much better feeling about the validity of the clicks that were charged to my accounts.
That might work if everyone had a static IP address, and if clickbot programs didn't exist.
Also, even the worst-case scenarios from vendors that market click-fraud detection services put total click fraud in the range of 14 and 20 percent for the PPC industry as a whole. As Trillanjedi said in another thread, "85.4% genuine traffic. Impressive statistic really, I can certainly work within that." And, of course, what matters in the final analysis is the advertiser's ROI, not the real or reported percentage of clicks that are invalid.
Would 85.4% be considered a reasonable confidence number in other industries (e.g. online payment processing)? PayPal's is 99.5% (so I hear). Also, there is some speculation that 85.4% is too high, due to the fact that fraudulent traffic can be made to look like ordinary traffic that doesn't convert.
Would 85.4% be considered a reasonable confidence number in other industries (e.g. online payment processing)? PayPal's is 99.5% (so I hear). Also, there is some speculation that 85.4% is too high, due to the fact that fraudulent traffic can be made to look like ordinary traffic that doesn't convert.
15 cents on the dollar going to waste? If its the cost of doing business, let google eat the cost, its their system that is faulty.
I think good evidence that Google does not effectively weed out fraudulent clicks is when you have keywords in your account which have a 200% CTR, that is one impression with two clicks.Unless I'm missing something, this suggests to me that one person and sit there and click on a keyword over and over and I get billed each time. There is no way of knowing if this is legitimate traffic or a competitor doing this.
Actually, this could be legit traffic. The clicks may have been processed before the impressions.
Would 85.4% be considered a reasonable confidence number in other industries (e.g. online payment processing)? PayPal's is 99.5% (so I hear).
I think you're comparing apples and oranges. A more sensible comparison would be to "waste circulation" (including circulation fraud) in offline media and direct marketing.
Still, why obsess over the things you can't measure when you can make intelligent decisions based on what you can measure: namely, ROI?
- If you're earning a good return on your PPC campaigns, then you probably trust the efforts that Google and other networks are making to combat click fraud.
- If you're losing money on your campaigns, then you'll have to decide whether to blame click fraud, your copywriter, or your skills in managing PPC campaigns.
- Finally, if distrust of Google and other PPC networks is keeping you awake at night, then why not find another way to drum up business?
Really, I think it would help A LOT.
Help do what? What exactly would they be doing. These numbers are simple a display of the actual clicks google dumps as fraudelent. There is nothing to verify. Nobody is getting charged for the clicks.
It is silly for adwords advertisers to be asking for a 3rd party vendor to do this. What exactly would they do?
Adsense publishers would have a better case as they are the ones that are potentially losing money from these stats.
A third party could watch google filter clicks all day long, it is not going to change the number. Google is not guessing at what the percentage is. This is an actual real number based on what the number of clicks discarded.
Feel free to tell me how a third party would change that.
I think you're comparing apples and oranges. A more sensible comparison would be to "waste circulation" (including circulation fraud) in offline media and direct marketing.
No, I'm actually talking about both systems' relative abilities to detect invalid transactions (since that is what both purport to do). There is already waste in PPC advertising which I imagine would not figure prominently in the number of invalid clicks (because people are unlikely to repeatedly click on ads they're not interested in, and such clicks aren't likely to come from "suspicious" sources).
Still, why obsess over the things you can't measure when you can make intelligent decisions based on what you can measure: namely, ROI?
It's just a matter of comparison of how well an engineering system does what it purports to do.
Let's consider what would happen if ROI of offline media increased because offline media dropped their prices, streamlined their business processes, etc. (This could happen.) What then would we say about a system that gives us 85.4% confidence that it is able to filter out invalid transactions? (Note I am not talking about Google, necessarily; I'm talking about PPC in general. This is one justification for changing to payment models whose engineering implementations can deliver better anti-fraud confidence.)
Huh? It doesn't happen that way. In an efficient advertising market click fraud merely devalues the bids for clicks.
For example, if I sell widgets and I can afford (according to my desired ROI) to pay $10 in advertising for each item sold. Let's say I know that in a perfect world (fraud free), my conversion rate is 25%. That means in a fraud free world I will bid up to $2.50 per click. But in a market where the fraud rate is 50% of clicks I still need to spend $10 in advertising for each item sold. My conversion rate is now at 12.5%. That means I will bid up to $1.25 per click.
Let's say the market allows me to sell 100 items a month:
Example 1
Revenue to Google = $2.50 x 400 clicks = $1,000
Example 2
Revenue to Google = $1.25 x 400 clicks = $1,000
The cost to the advertiser is the same too. What really hurts is that as a publisher with legitimate traffic (versus cheaters) we get less than our fair share. That's because we are delivering clicks worth $2.50, while the fraud sites are delivering clicks worth $0.00.
The only thing that can send your ROI to the toilet is competition.
Huh? It doesn't happen that way. In an efficient advertising market click fraud merely devalues the bids for clicks.
publisher math ;)
No, just economics 101
My campaigns with the highest CPC have the highest percentage of invalid clicks.
I *thought* mine would be this way, but it turned out not to be case.
Don't have any accounts at the moment with more than 5% invalid clicks (as accounted by Google) and even if I allow for maybe another 10 to 20% that Google isn't catching, we're still doing acceptably well. Of course I'd like it if there were better ways to detect and prevent the invalid clicks, but as long as the bottom line remains black, I can maintain my equanimity.
In a previous life I worked for years in catalog direct marketing - do you know the response rate on THOSE? I haven't kept up, but at the time it was considered to be doing quite well if we hit a 5% response rate on say, 100,000 catalogs mailed out. Compared to that, this is gravy.
Google should also include the IP address of clickers to your site in a report
they would allow the advertisers to view the IP addresses of the clicks
So the plan is to ban clickbots that Google has already identified? What's the point of that?
there needs to be a 3rd party that audits this
This is all heading to 3rd party verification
ROI is the perfect impassionate auditor of click quality. In fact, I wouldn't care right now if 95% of my clicks were "invalid" as my ROI is still excellent.
Let's not forget that we're all in this CPC/contextual text advertising game because of how much more efficient it is than other forms of advertising.
I'm sure the click fraud report got some attention from AGs and I expect some REAL investigations to commence very soon.
[edited by: TypicalSurfer at 4:12 pm (utc) on July 27, 2006]
