Digital advertising continues to show tremendous growth and so does the ad fraud in this sector. Ad fraud refers to an impression-based fraud which is an attempt by cyber criminals to serve ads that have no value for the advertiser. An advertiser would get charged for creating an advertising impression which will never be viewed by a real human. This non-human traffic on the websites leads to a huge wastage of advertisers’ funds and time.

The latest statistics suggest that more than $11 billion was wasted in 2014 on serving ads fraudulently. This digital ad fraud is an industry growing at an alarming rate of 22 per cent year-on-year and needs to be checked.

Fraudsters function in several ways but the most common and prevalent methods are via bot traffic, ad stacking and pixel stuffing.

Design behind the deceit Robots or bots are automated pieces of codes crawling the web and are the biggest contributors to this problem. They are becoming highly advanced and their latest breed, also referred to as humanoids, are so sophisticated that they can actually imitate the browsing behaviour of humans, hence becoming increasingly difficult to track.

Display banners used to be the focus of these bots but with higher budgets being shifted to videos and mobile, the trend is shifting. As per the latest trends, bots account for 11 per cent of display ad views and 23 per cent of video ads. And these bots are not restricted to any specific publisher or network. They are prevalent across all categories of publishers and across all networks, though programmatically bought ad impressions show higher percentage of bot traffic than individual publishers.

Apart from these bots, ad stacking and pixel stuffing are also quite rampant. Ad stacking is the process of placing multiple ads on top of each other in a single ad slot. With ad stacking, only the ad on the top is viewable, but the ads underneath are not visible to the end user though they are counted as impressions.

Pixel stuffing is the process of serving one or multiple ads in a single 1x1 pixel frame. Again, since these ads are invisible to the naked eye, money spent on such ads is essentially money dumped.

Mounting a defence It is clear how big a worry ad fraud is for the advertisers as it has a serious effect on their return on investment (ROI). “Be that programmatic or however our ads are delivered to consumers, we should get what we pay for and that our ad is viewed by a consumer. Industry standards are just not adequate enough for our requirements,” says Sarah Mansfield, VP, Global Media, Hindustan Unilever Ltd. As a matter of fact, a long list of marketers such as Procter & Gamble, Mondelez International and Coca-Cola are reviewing their media-buying business as they look to cut costs and find agencies skilled enough to address the issue of ad fraud.

Fortunately, much to their relief, a new breed of ad tech companies focused on fighting digital ad fraud have developed algorithms, data mining and analytics to act as a safeguard against ad fraud.

These continuously evolving analytics engines essentially work with a huge amount of data — definitive and statistical. Definitive data, as the name suggests, is a high-quality premium data which ensures that the request is from a human — for example, the data obtained from users who have successfully solved a captcha. In addition to this definitive data is statistical and probabilistic data which is derived from real-time analysis and audience profiling based on multiple parameters such as browsing behaviour, Java Script calls, tracking the pointer movements of the mouse, IP quality and frequency of requests.

Reclaiming the ecosystem These latest technologies are a boon not only for the advertisers but for the entire digital ecosystem. They also help the publishers against these malicious bots, leading to hefty savings on their precious tech infrastructure, safeguarding their content and providing a much better experience to the end user.

At the same time, these tech platforms also help notify the innocent users whose computers have been infected with these bot malwares, help them recover their machines to run on their optimal power and save on the bandwidth cost that is otherwise consumed by these bots to access hundreds of websites in the background.

What we need to understand is that even for a serious issue such as ad fraud, there is no platform which can claim to filter bot traffic with 100 per cent success. It is a continuous and ever evolving fight between the good and the bad with both the parties trying to stay ahead of each other.

To be able to make these cyber gatekeepers much more effective, there needs to be greater transparency and all the related parties — publishers, advertisers, media agencies and the ad tech companies — need to work closely with each other to handle this threat.

Amit Mittal is Co-Founder, Simpli5d Technologies