when a user clicks a sponsored link, the advertiser pays, maybe 10 cents, in some cases on the order of a dollar. A user hostile to the particular advertiser, or to an entire category of advertisers, could thus click on their links and have them waste money. Maybe good for Google (they get the money) but bad for the overall business model.
Has this issue ever come up? Does Google do or claim to do anything to detect and counteract such activity?
They track the ip of the click and pay by the individual IP’s clicking the ad, not everytime the mouse button is hit.
This is called click fraud; and yes, it’s a big issue in the online advertising world.
Wikipedia has an OK article on it:
I don’t know what specific countermeasures Google uses, but I’m sure they do something, or else advertisers would take their business elsewhere. But then, Google is the proverbial 800-pound-gorilla of internet advertising, so they have a lot of leverage.
Click fraud is notoriously difficult to define, let alone police, so there’s no clear-cut answer really. It’s not going to be quite as simple as “per click per ip”, though.
The method suggested by pan1 is insufficient in some ways, and overbearing in others. On the insufficient side, you have to deal with botnets – thousands, and in some cases millions of infected (almost entirely Windows) machines that can be directed to perform all sorts of distributed activities, including click fraud. A per-click-per-ip scheme is still quite abusable by botnets.
On the flip side, per-click-per-ip scheme will also wind up blocking a lot of non-fraudulent clicks because of NAT. Lots of homes, offices, libraries, etc. have multiple users behind a single IP address. I don’t know what the actual statistics are, but with the continued crowding of the IPv4 namespace, it’s not at all unrealistic to expect that there are probably several hosts per allocated IPv4 address (on average). That’s a lot of potential legitimate ad revenue to just throw away on general principle.
As another example of IP sharing, ISPs may reassign the same IP. I remember back in my AOL days I’d sometimes be refused at some web sites (polls, for example) because my IP had already been used by another AOL user earlier in the day. The best I could hope for was to sign off, reconnect and hope for a new IP that hadn’t already been used at that site.
Yes, and that can still happen today. Every time I reboot my router at home, it winds up with a different WAN IP address. Apparently Verizon uses short leases with no reservations 
Similar issues can also arise from the use of proxy servers or Tor. Early on, at least, Tor had issues with Wikipedia slapping IP bans on their exit nodes because someone was using the network to vandalize wiki pages. The same kind of thing can happen when some jackass uses Tor for IRC abuse and the entire exit node gets k-lined.
There are, in general, a lot of problems with trying to use an IP address as a (sole) identity mechanism., or with assuming that there is a 1:1 mapping of internet users to IP addresses.
I’m a roofer. I can expect clicks to cost minimally $10 per click. Ouch.
Last time I priced Virus cleanup it was like $6, unfortunately a virus cleanup does not go for anywhere near the price of a roof.
hmmm… a new computer (usually) costs less than a new roof, so that kinda makes sense… :dubious: