gsmith
msg:3660691 | 10:04 am on May 28, 2008 (gmt 0) |
It looks to me like the fraud claim rests here (from the same article): | Click fraud can be detected by software that can track suspicious patterns, such as repeated clicks from the same source. Although Citysearch assures its customers that it applies this technology, the experiences of many of its customers shows otherwise, according to the suit. Still, customers are led to believe that Citysearch is in fact actively fighting against click fraud. |
|
|
civgroup
msg:3660979 | 3:37 pm on May 28, 2008 (gmt 0) |
We quit advertising with them over this - it was clear to us that click-fraud was going on, but they did nothing when we reported it. I'm glad someone is holding their feet to the fire over it.
|
jimbeetle
msg:3661029 | 4:31 pm on May 28, 2008 (gmt 0) |
| Citysearch...pays commissions to its salespeople based on the number of clicks their customers' ads receive, providing an incentive for click fraud, according to the lawsuit. |
|
Can't really think of another way to base a salespersons commission on other than account revenue. Are there other models used in PPC sales?
|
idolw
msg:3661031 | 4:34 pm on May 28, 2008 (gmt 0) |
| Can't really think of another way to base a salespersons commission on other than account revenue. Are there other models used in PPC sales? |
|
I guess it is not about the model but the rules within the model. I guess Google/Yahoo!/etc. do not offer incentives for reps to click on clients' ads.
|
Rodney
msg:3661081 | 5:19 pm on May 28, 2008 (gmt 0) |
I'm probably missing something pretty obvious here, but how can the salesmen increase the clicks on a customers ads? Seems like that's something out of their control.
|
jimbeetle
msg:3661102 | 5:29 pm on May 28, 2008 (gmt 0) |
| but how can the salesmen increase the clicks on a customers ads? |
|
By clicking on them?
|
whoisgregg
msg:3661121 | 5:47 pm on May 28, 2008 (gmt 0) |
| but how can the salesmen increase the clicks on a customers ads? |
|
By clicking on them? |
|
And...
Asking their friends and family to click on them.
Hiring one those "100,000 visitors to your site" services.
Running a click bot.
|
jimbeetle
msg:3661133 | 5:59 pm on May 28, 2008 (gmt 0) |
| Hiring one those "100,000 visitors to your site" services. |
|
Or owning one. Still, what other models of salesperson comp are there in the PPC arena?
|
LifeinAsia
msg:3661152 | 6:10 pm on May 28, 2008 (gmt 0) |
| Citysearch...pays commissions to its salespeople based on the number of clicks their customers' ads receive, providing an incentive for click fraud, according to the lawsuit. |
|
Can't really think of another way to base a salespersons commission on other than account revenue. Are there other models used in PPC sales? |
|
Well, the obvious would be commissions on non-fraudulent click revenue. In other words, implement fraud detection measures and only bill the customers and pay commissions on non-fraudulent clicks.
|
kapow
msg:3662199 | 7:18 pm on May 29, 2008 (gmt 0) |
There are only so many things you can detect about a visitor (click):
- IP
- Client (browser)
- And other variables sent by the client (OS, colours, resolution... etc).
IP is the hard one, but if you can fake the above, AND create a 'relatively random pattern' algorithm, Voila! Your click-bot is undetectable.
|
Receptional Andy
msg:3662236 | 7:38 pm on May 29, 2008 (gmt 0) |
| IP is the hard one, but if you can fake the above |
|
You can't fake an IP with a clickbot: TCP/IP just doesn't work that way. You can only use IPs under your control, and to produce pseudo-random results you'd need quite a few (i.e. a botnet) to avoid easy detection.
|
StoutFiles
msg:3662239 | 7:43 pm on May 29, 2008 (gmt 0) |
IP's are given out in blocks; you could get plenty of them if you take the thousands of extras that corporations and/or universities have. However, I'm sure enough IP's in the same block would be easily marked as click fraud.
|
|
