Click Fraud = Friction for Google

Fraudulent web click are friction in Google’s business. Fraudulent clicks ad costs to the system without a benefit to the performance of the system. Google’s profit is derived from improving the system (of finding customers for advertisers) and taking a cut of the profits that their system creates. Google makes a great deal of money because their system of matching advertisers with dollars to spend to customers. Google does this through ads on their search results pages and on third party web sites. Google engineers will do whatever they can to find ingenious ways to reduce that friction.

There have been many stories over the last few years about click-fraud. But none I have seen explain the simple idea that Google is the company with the most to gain by eliminating it (and Yahoo next). They often point to companies suing Google about fraudulent clicks instead.

Companies like Google run ads on web sites and charge the advertisers for each click (anywhere from a few pennies to several dollars for each click). Advertisers want to get potential customers when someone clicks on their ads, obviously they don’t want to pay when there is no chance the “visitor” is going to become a customer. Obviously, fraudulent clicks are bad and those that engage and encourage such behavior are acting unethically and immorally and should be stopped and punished.

Now, Lets look at other advertising. Advertisers pay large (say tens of thousands of dollars or more) for ads in magazines and on TV. They base those ads on guesses about how many potential customers they may reach and how many will actually look at the ad. I think this is important to understand to set the context of measuring advertising success. Advertisers normally have to rely on making tenuous connections between ad spending and benefit to the company.

With web based advertising, companies can much more easily measure the effectiveness of ads (especially for immediate responses – “brand building” can also be measured more effectively but not quite as easily as immediate purchases). So If an advertiser choose to pay Google $30,000 for ads last quarter they can get a very good idea of what the payback on those ads were (especially compared to traditional advertising). They can decide whether it was worthwhile or not.

That decision can be made irrespective of fraudulent clicks. To make things easy lets say the company says they received a benefit of $40,000. Again to make things easy lets say the $30,000 figure and $40,000 factor in all other costs and expected returns… That is the company has a net gain of $10,000. That means they are likely to want to continue such advertising.

Now lets look at the fraud. If there is fraud involved that the $30,000 figure has been “inflated.” The $40,000 is the benefit (regardless of how much fraud happened to be involved). Well lets look at some options, if the company measured the return at $20,000 instead of $40,000 they would decide to stop running ads since the cost was greater than the benefit. Again we are making the numbers represent all factors (in an real example, a company might expect that they improve how they use the advertising over time so even if they took a loss might see it as worth continuing for the future). So in this case the $20,000 return factors in all benefits to the company.

So they stop running ads; what happened. Again to make things easy lets say Google makes a 40% profit, the fraudsters make 10% of the total click cost, legitimate web sites make 30% (given the $20,000 benefit the company saw that some clicks were legitimate), and 20% is taken in costs of running the system. Google made $12,000 (40% of $30,000) the fraudster made $3,000, legitimate web sites running Google ads make $9,000 and the company lost $10,000. The next quarter no one makes anything – they run no ads. These numbers are just made up but the logic is not dependant on what percentages are chosen.

So future ads would not be placed as the company did not see a return. Google does make some money from the short term fraud. As the market moves to equilibrium however advertisers will run ads when they make money and not when they don’t. For the sake of this argument we only look at ads on non-Google sites that run Google ads.

Google is able to make money by serving as the intermediary between those wanting to run ads and those wanted to buy ads. Those wanted to buy ads will pay Google for ads that pay them, they won’t pay Google if ads don’t pay them back. So assuming zero fraud, Google does very well, the web sites running ads do well and advertisers do well (they chose to run ads that make them money and not others). When the advertisers are making a great deal of profit they will buy more ads. For example, they might increase the price they are willing to pay to get there ads placed on more web sites (Google’s system will give more prominence to ads that pay higher amounts all other things being equal) and add news ads about related topics (since the ads they have so far are so profitable).

If fraud increased the costs for advertisers, Google would get a portion of that money. So I can see why people could look at this and say Google benefits from web click fraud. But this is only true in a very limited view. Google already has the advertiser committing to run ads. They will keep that advertiser paying them money as long as the advertiser is making money. It is in Google’s interest for the advertiser to make money so they will increase the amount of money they pay Google. It is true some poorly run companies focused on short term gains and without ethics (there sure are plenty of such example) would kill their long term viability for a boost to earnings today that fraudulent clicks would provide (not that the company would encourage it, but they would fail to take vigorous action to prevent long term damage to their business to try and gain huge bonuses… by short term effect of not preventing click fraud).

However that is a very stupid business strategy and not one Google would adopt. You don’t even have to assume they are ethical. It would just be far too dumb for Google’s long term interest and number focused culture to proceed on such a path.

At this time I am very confident Google is spending huge amounts of money to combat fraud. Because that fraud will cost them much more in the long term than it costs advertisers in the short term. If the fraudsters can successfully steal money that would otherwise go to sites Google places ads on (that in turn provided those advertisers potential customers) that will be a huge loss of potential business for Google.

On top of that, my belief is the people at Google just think the fraudsters are bad people. And they love to use technology to prevent people they think of as bad from taking money from good people. Engineers in general love to work on making things work well. They don’t like inefficiencies and friction in systems. Fraudulent web click are friction in Google’s business. Google engineers will do whatever they can to find ingenious ways to reduce that friction.

I am a bit confused why Google doesn’t adjust the contract to state that the charges are per click, Google does what it can to prevent false clicks and that is the way it is. I don’t really understand how these law suits have any merit. If Google were being fraudulent ok, then yes but…

If they are paying $1 a click and making 90 cents then they cut what they will pay and Google losses. If they then say that well that isn’t fair if I pay only 90 cents per click my ads don’t make it because others pay more well that is not a Google issue. Unless they claim they are subject to special fraud that other advertisers don’t have. This can then get into the “clicking on competitors” ad fraud that is mentioned in some of the articles. I hope that this is illegal (fraud is illegal in whatever form it takes…). And so hopefully the people that engaged in such behavior would be fined and put in jail. And while it is different, it is also just friction to the system. To the extent that exists the system functions less well and therefore there is less profit from which Google can take its share. So again logic dictates Google would do everything they can to prevent, in the short and long term, such behavior.

“I would have written a shorter [post], but I did not have the time.”
– Blaise Pascal

Tags: google adsense – adwords