Online advertising is a great idea that often has some troubles in execution. While other forms of advertising are sold at flat rates—using the amount of space used or the amount of time it runs, sometimes both, as the metrics to determine pricing—online advertising is a much different animal that runs on several different metrics. Traffic generated is easily one of the biggest such measures, but even that brings with it the potential for fraud in the form of “non-human traffic,” a development that Chinese advertisers are finding has a lot more impact than some might think.

Indeed, a new report from Miaozhen Systems shows that non-human traffic caused roughly $1.6 billion in losses to advertisers. With Internet use rapidly on the rise in China—at the end of June, the country accounted for fully 591 million users, and mobile users came in at 464 million--many companies were interested in offering up advertising to the population. Indeed, the online advertising market in China is set to break $20 billion this year, proving there's plenty of interest in offering products and services to somewhere around 600 million people.

But non-human traffic has always been a problem for advertisers online, and the rise of fraudulent impressions and clicks has posed both a loss to advertisers and in sheer credibility to the online advertising concept. But the issues in the Chinese market don't seem limited to issues of machines clicking on ads where humans should be; other issues are also being raised in the Chinese online advertising market.

One of the biggest such issues is geography. Publishers are sometimes seen routing ads to sparsely-populated areas when advertisers paid for access to “first-tier cities”; indeed, at last report, 22.12 percent of campaigns ran into this particular issue. Timing was also a problem, as those who paid for prime time access were foisted onto the midnight-8 am slot in 14.81 percent of cases. Finally, issues of exposure were also a factor, as 4.2 percent of campaigns were found shuffled onto less popular matter than which the campaign originally called for. The rise of mobile use in China has prompted likewise shifts onto mobile devices for advertising, and even these were found to fall prey to similar measures. Mobile publishers were spotted moving advertising to inferior apps, displaying advertising on non-iOS platforms when iOS was more desirable, and more standard traffic cheating measures like registered users were also in play.

New standards, like API and SDK, are set to offer greater supervision and help bolster healthy development, but these may well not be the ideal means. There's one axiomatic observation here that describes much of the computing world: what one programmer can do, another can undo. This particular axiom has applications from copy protection to traffic fraud, and many users have discovered that, just because a system—any system, really—is protected today doesn't mean it will be protected tomorrow. 

A simpler solution—potentially a better solution—might be to realize this issue and, instead of establishing complex formulae to online advertising sales, sell online advertising in the same way that other media is sold, based on total viewership reach and duration of the ad. Cheating that is a much tougher prospect, since it is inherently simpler; any system with fewer moving parts has fewer points at which it can break down. This may not be the perfect solution, but it may well help address many common issues in online advertising.