Digital advertising is considered to be the most efficient and transparent medium for advertisers but certain recent statistics make one question how true this claim is.
A December 2014 joint report by the Association of National Advertisers (ANA), a representative body for US marketers, and White Ops, a US-based digital advertising security company, states that global advertisers will lose $6.3 billion to bots in 2015.
The study, which saw participation of 36 ANA members, analyzed digital marketing campaigns for a period of 60 days. It found that bots account for 11 per cent of all display ads, 23 per cent of video impressions, 17 per cent of programmatic inventory and 19 per cent of retargeted ads. This does not include the non bot-related frauds that digital advertisers have to also contend with, where media agencies, networks or publishers might charge for non-genuine traffic.
In a more recent study, mobile ad tech firm AppLift found that 34 per cent of all mobile traffic is at risk of fraud. Even the IAB, in a joint report with EY released this month, said that fraudulent impressions, infringed content, and malvertising cost the U.S. digital marketing, advertising, and media industry $8.2 billion annually.
That the rosy picture that we perceive of digital advertising might not be entirely as it seems was brought to attention in dramatic manner when Google said that 56.1 per cent of all ads on its network are not viewed by the audience. This means that advertisers have been paying for ads that were not even seen by their target audience. A few other industry observers had made similar claims earlier but it took the admission of the world’s largest content network to actually start getting people to take attention.
Here it is important to note that lack of viewability might not necessarily be due to any dishonest activities on the side of the ad networks or the publisher. The reasons for low visibility are more likely due to slow networks and other technical problems.
However, there have also been various instances where publishers resort to a tactic called ad stacking; placing multiple ads on the same ad spot, to boost impressions even though the ads are not actually seen.
Conversations with planners and digital executives reveal that a shady underbelly does exist. Anecdotes about clients requesting high traffic, even if it included bots or remaining unconcerned about the overall quality as long as the campaign is within budget are common. Even certain agencies and ad networks are not above passing off (and charging for) fake Likes, followers, clicks or impressions to the client.
That fraud exists should not be surprising. But it seems that marketing heads (and others) are either not aware or turning a blind eye towards these issues.
Many a time media planners, publishers and ad networks themselves are victims to external malignants (bot networks, hackers, etc). Unfortunately, there is no study available for how much Indian advertisers lose because of these activities, whether done on purpose or occurring without anyone’s knowledge. This brings us to the other question; just how big is this a problem for India?
With advances in technology, the magnitude of the problem is becoming more apparent. “"Ad fraud has existed all along, but its shape and form has evolved over years, just as the fraud detection tools and awareness in the ecosystem around ad fraud has evolved too. Impression level fraud has been on the rise in the last few years with robotic impressions, domain masking and ad stacks being some of the latest challenges. Thankfully, we have proprietary algorithms as well as third party tools available today to combat these challenges and detect and block fraudulent inventory”, opined Satish Kadu, Founder & CEO of YOptima, programmatic audience buying platform.
“There are two things that are happening,” adds Swapnil Shrivastav, Head (Ad Tech) at Times Internet. “Fraud is growing and ad money is shifting to CPA (cost per acquisition) model. The second is because of the first but due to a CPA model, no one seems really concerned. But at the back of everyone’s mind, they know that this (online fraud) is common.”
When asked whether publishers are aware of ‘fake’ traffic on their properties, Shrivastav pointed out that to actually track the traffic quality requires additional algorithms to be added to the website, which only those publishers with their own ad stacks will know how to do (Times Internet has started its own ad tech platform to reduce dependency on external networks, said Shrivastav). “The publisher is usually in the dark as there is no in-built system in ad servers. Frauds mostly occur when people buy on open RTB where it can go as high as 30 per cent. When people buy from premium publishers or closed networks it is less but can be as high as 20-40 per cent for long tale publishers,” he said.
On the other hand, the head of an ad tech firm told us that publishers also tend to resort to immoral methods to boost revenues. “We have known publishers that purchase traffic, which is sometimes bot. They are aware of it but still monetize it. It does tend to happen,” he stated.
Kadu is of the opinion that on the publisher side, fraud is more likely with inventory that’s resold via a daisy chain of ad networks, as many publishers may not have enough visibility on how or who is selling this inventory. “We don’t think that premium publishers would risk their brand image by indulging in such activities, but some long tail publishers and affiliates might be tempted to earn quick money by engaging in or supporting fraud”, he said. There is also the fact that with most premium publishers, the majority of the inventory is directly sold rather than through RTB or programmatic, thus reducing chances of fraud.
Lavin Punjabi, MD of Affinity, says that fraud or bot activity is prevalent on all kinds of online media - display, social, video, search, email, and affiliate marketing. “My gut tells me that it’s between 5-10 per cent on any kind of media,” he told us. He agreed that if the client asks for cost per conversion model than it puts more pressure on the ad network to put in more controls and be more careful.
Most industry experts we spoke with seemed to agree that though digital fraud is a reality (and seems to be growing), it can be combated by working with trusted networks and paying attention to where exactly the digital spends are going. For example, Karan Gupta, MD of Andbeyond.media, opined that the fact that India is not a huge CPM market, insulates us somewhat from major botnet operators. Though this situation might change if digital advertising continues to grow as it has been in recent years.
Add to this the fact that most clients and agencies work with Google and Facebook, which have strong anti-fraud measures built in. Though, as the recent hacking of Yahoo’s ad network earlier this year along with similar attacks on Google, New York Times and Reuters showed, not even the big networks are completely immune. But there are other third party tools available to protect from bots and help track ad performance
“If your (client’s) digital strategy is flawed then you will get duped,” says Punjabi. For clients, Punjabi advises getting a centralized platform with third party vendors like Moat or IAS and to work with trusted networks. But working with third parties adds another cost layer and this is something that most clients do not want. “If you are spending Rs 100 crore on advertising and you are losing 10 per cent to online fraud or bot traffic, why can’t you spend just a per cent of that cost to protect yourself?” he argues.
Kadu also advised users to request for transparency from their network partners. “Clients have been used to working blind. The more they ask for data, the more fraud will be detected and this will act as a deterrent,” he said.
But it is just not about data, many of the ad tech professionals we spoke with pointed out that advertisers need to start showing more respect to media inventory and be willing to pay more for better quality traffic instead of cutting corners.
“After the arrival of programmatic advertising with world-class ad solutions in India who automatically detect and block bot traffic, fraud is no longer a paint point. The bigger concern today is the growth of ad blocking which is ruining publisher revenue,” said Reem Saied, Business Head (Cadreon India) for IPG Mediabrands. He, however, agreed that if clients were not careful and continued to buy through affiliate networks via RTB, rather than trusted networks, then they would be susceptible to such frauds.
The very fact that traditional ad networks are dying out, at least in India, is also something that is aiding the fight against bots and fraud traffic. As the chain between publisher and advertiser gets shorter, the chances of any one component in the chain straying from the straight and narrow is automatically gets reduced.
With most clients now concentrating their spends on premium publishers and networks, the chances of fraudulent traffic also decreases due to more stringent protection. As mentioned earlier, that the digital ecosystem in India sees unethical practices should not be a surprise; as this seems to be the case even the more developed markets. That people, especially clients, are turning a blind eye to it might be more of a cause for concern in the long run when the ecosystem actually matures and more money is injected into online ad campaigns on a regular basis.