
In December 2014 the revealed that advertisers will lose approximately $6.3 billion globally advertising to bots in 2015. A 57-page report highlighted the alarming severity of the problem that has penetrated the online advertising supply chain as a whole. And it raised a more fundamental question about who is actually viewing your digital advertising in the first place, and whether or not it is even a human being?
Cybercriminals are opportunists following the money, they have caught-on the quality versus quantity game and are attracted primarily to higher CPM campaigns. As such, it is to no surprise that the more lucrative video is a sitting duck: 23% of all video impressions are fraudulent. And bot traffic in premium programmatic inventory averages 17%, while 19% of retargeted ads are compromised. This leaves lower CPM campaigns open to attack by less sophisticated bots that look to cheaper or performance-based buys, they have colonised social and currently account for 11% of display impressions.
Ad fraud has grown to its current scale partly because of how long it has been left unchecked. While it has been widely reported in the press that up to 50% of online traffic is fraudulent, digital ad budgets continue to grow (by 14% YOY according to the IAB’s latest 2015 report) with little evidence of budget allocated to protect ads against the threat of bots and to ensure efficient use of digital spend.
No doubt there is an element of fear in this delayed reaction, but marketers and the industry at large can and must begin eliminating bot fraud by implementing an action plan - today. The ANA and White Ops report put forward a robust list of 17 recommendations to resolve our online vulnerabilities but, in truth a huge difference could be made if advertisers would follow three simple steps as a starting point to protect their digital campaigns from ad fraud.
1. Monitor all traffic with a consistent third party tool – comparability over time is essential
This should validate or disprove assumptions about the quality of a publisher or an ad tech company’s traffic, despite premium classifications. In particular, buyers should seek transparency over sourced traffic, any method by which publishers acquire more visitors through third parties, as traffic sourcing correlates strongly to high bot percentages.
2. Update blacklists frequently and narrowly, as often as daily, and control for ad injection
Ad injection is the unauthorized placing of ads on sites where they do not belong. It can cause Programmatic buys to contain higher levels of fraud, so buyers should discuss this with their DSP or tech platform, to understand the problem and insist on transparent solutions.
3. Concentrate advertising during audience waking hours
The majority of fraudulent traffic comes from everyday computers whose browsers have been hacked. Bots not only blend in with human traffic, but are being actively targeted, unbeknownst to advertisers. Fraud activity peaks when ‘real’ users are sleeping, but their computers are still awake, between midnight and 7am. Additionally, older browsers like IE6 and IE7 are reportedly more popular with bots than humans, so it is advisable to reduce buys on these dated and vulnerable systems.
The industry-wide preoccupation with viewability has detracted attention from the alarming extent of online ad fraud. After all, standard technical measures of viewability do not ensure humanity, and while advertising focuses on views, it altogether ignores the nature (literally!) of the audience it is attracting.
Agencies, brands and publishers must take the responsibility to develop an understanding of this issue as it poses a genuine threat to the sustainability of the advertising ecosystem, as we now know it. Players that authorize third-party traffic validation technology, and implement the recommendations set out above to immediate effect will be the pioneers in this space. The winners, however, will be the brands, as they will finally see their digital advertising budgets put to efficient and effective use – advertising to humans, not bots.
Recommended