Click Fraud, Google and simple IP solution

I don't understand Google when they allow one IP address to register multiple clicks on a single Google Ad on any given day. Surely, the assumption is it's the same person, given the IP address is the same? Surely advertisers should be paying for unique hits, not simply 'clicks'! 10 clicks from one malicious person is the same cost as 10 unique visitors with Google Ads. This is crazy. I realise it MAY not be the same person if they are on a network, but cut the advertiser some slack and prevent click fraud by simply not registering a 2nd and 3rd (etc) click from the same IP address on the same advert on the same day. This is really simple to code.

The reason is 'proxies' I work for an American company but my office is located in the UK, but everytime I go on the Internet I appear to be in the US.

Why? Because all Internet traffic from all offices globally (that's about 40 countries) gets trunked to the US on the WAN links and then goes out on the net through the firewall in the US. And here's the kicker - all those guys from Hong Kong to St. Petersburg to Rio all use the same US IP address or at least just a small bunch of addresses. Many companies work like this. IP address is by no means a foolproof way to count genuine clicks.

If you use a custom tracking solution that gives you keyword history (weighted clicks) you'll notice that many users click on several ads for different keywords before making a purchase. An example would be:

Keyword Time Date
holidays in spain 14h00 31/08/06 No Sale
hotels spain 14h30 31/08/06 No Sale
expedia 21h30 31/08/06 Sale

In a case like this, it wouldn't be fair to Google if the advertiser only pays for the first click.

There are many programs that will bounce you off of 100's or even 1000's of servers to mask your IP.