Newsweek On Click Fraud, Search Engine Response

prostoalex writes "Newsweek magazine says click fraud is the bane of the search advertising industry. Google and Yahoo! are apparently working on the standardized definition of a "good-faith" click in order to weed out the fraudulent ones. Meanwhile, merchants like Assaf Nehoray are taking their money elsewhere, getting abundant clicks, but no real revenue on Internet advertising campaigns. Newsweek also mentions Google suing a Texas company for placing the AdSense code and then clicking on it in order to run up the revenue. John Battelle says that his friends in the search industry tell him the click fraud is growing and that changes are not too far away."

Comments

You're about six months behind

[Anonymous Coward]

Google's Fraud Squad Battles Phantom Clicks

Click fraud?

[mistersooreams]

What exactly is this click fraud thing? I can't really see how it can be exactly defined. Maybe the owners of the website occasionally want to click on their own adverts because (*shock*) the product is actually relevant to their site, and thus to them. In fact, relevance is supposed to be the whole idea of Google's TextAds, isn't it?

Obviously someone genuinely wanting to click their own ads ten thousand times is rather unlikely, but where do you draw the line? Is this written in a contract anywhere? What about getting other people to click the ads for you?

This seems to be a very fuzzy legal matter. I'm as pro-Google as the next Slashdotter but I can't see how they have a water-tight case here. That said, I'm not an expert, so perhaps someone can correct me.

Re:Puff

[iconnor]

Is there any service that advertising can pool their useless clicks?

To an advertiser, a useless click is a click that hits the adwords landing page and little (or nothing else) and does not mean a sale. If these bogus clicks could then be processed at a 3rd party auditing house, then fraud could be detected and each member could then complain to google about bogus clicks.

I am sure someone must have thought of this already - I just can't find it listed on google :(

Re:Uh duh...

[fyngyrz]

Disagree. Pay per click isn't the problem. Bad (misleading) ads are the problem; there's no reason to presume that pay per click will die because of these. The advertisers will die, instead. And where's the downside of that?

There are perfectly reasonable ways to use Google ads, for instance. Describe your product honestly, market your product honestly on the target page the ad leads to, and provide a good and well supported product.

If you can't be bothered to do that, then you deserve to have your ad budget eat you for dinner, IMHO.

If you do follow those basic guidelines, then the ads will bring potential customers by your pages, and some percentage of them will actually purchase your product(s) and/or service(s).

There's no magic to this -- unless you're a fraud right out of the gate anyway.

That's not to say that you can't work the system in such a way as to make it legitimately benefit you. For instance:

Some of my competitors do such a poor job, I decided to put Google ads right on our sales pages so our customers could easily find our competitors. The results they get are so yucky that I consider them to be marketing for us, in a reverse sort of way. There's nothing on our pages that say "go look at how crappy our competitors are, click the ads and then come back" or anything like that... the ads just sit there, advertising software similar to ours... and our sales picked up about 20% over four weeks once we put the ads on. Apparently, our customers are savvy enough to know where they've been -- and how to come back -- when they wander off to look at these other folks. And the funny thing is that we get paid for all this. Now, if our competitors are silly enough to keep advertising a shoddy product, why, I'm simply delighed to host their ads. :)

Re:3rd world countries

[britneys 9th husband]

I think you're on to something with the "hit the monkey" model. If you think about it, Google click fraud has a lot in common with crapflooding Slashdot. In both cases, you have a malicious user trying to do something (e.g. post a comment or click on an ad) far more times in a short period of time than any legitimate user ever would. It becomes a cat-and-mouse game. The difference is that with Google ads, the stakes are much higher, and the techniques used (both by the malicious users and the site trying to stop them) are more sophisticated.

An obvious thing to look for is lots of traffic coming from the same IP or subnet. That's why Slashdot has IP bans, makes you wait 2 minutes before posting another comment, etc. Google of course can look for similar patterns. Therefore malicious users need to make their traffic look like it's coming from all over the place instead of just one computer. The GNAA used to crapflood Slashdot by compiling lists of hundreds of open proxies and writing a script to have them post comments all at once (Slashdot no longer allows open proxies to post). We can assume Google filters out ad clicks from anonymous proxies. So now the malicious users need a way to recruit hundreds of computers, preferably from lots of different subnets, and without using open proxies. You can do this by paying people to click ads from their computers, and "hit the monkey" is probably the cheapest known way to do this. If you're making more money on clicks than what you're paying your army of clickers, you'll make a profit.

Now, I suppose you could also write an automated program to do the same thing, but that would be called a "virus" or a "worm," and these things tend to attract a lot of attention from various law enforcement agencies. Better to pay people a couple cents to hit the monkey than go to prison.