Digital advertising fraud concerns from media buyer’s aspect


Campaign renewal decisions can be really hard to make with commonly used analytical tools. Especially nowadays,  where fraudsters spread the industry. Media buying is all about testing and optimizing traffic segments but at a certain point, it’s hard to indentify the reason of some sources’ low performance. Google Analytics – the industry standard for traffic analysis – is a great platform to track and optimize campaigns, but when spotting fraud is in the focus it has it’slimitations. Every fraudster knows how to fly under the radar, do a quick hit ’n’ run and monetize, so based on the bounce rate, time on site and pageviews/visits it’s sometimes quite a challenge to spot weak elements in your portfolio.

Botnets are perfectly tailored to dodge the attention of the decision maker – who works based on the given metrics. They are capable of attempting multiple pageviews, they sign up, so your conversion and CPL meter won’t ring the alarm. Then you realize that your user base consists of a bunch of fake users and they aren’t spending a dime.
No matter what adserver you are utilizing, the available metrics won’t highlight all fraudulent activities. Neither will your internal admin statistics or GA, whichever you are working with.

Another important topic is that programmatic advertising is rising. According to a report released by IAB and Pricewaterhouse Coopers $10.1 billion was spent on programmatic display ads in 2014, and a recent Advertiser Perceptions survey reports that 78% of high-level decision makers are now using programmatic technologies and strategies across campaigns.

Programmatic buying allows media buyers to buy “eye-balls” – that is, only buying impressions that are going to be seen by a qualified viewer.

Or so the advertiser hopes.

The concerns around programmatic buying are largely focused on fraud and viewability – is the impression being served to an actual human? And if so, do they actually have the potential to see the ad?

In a May 2015 survey, fraud and quality/brand safety concerns were the leading obstacle to buy ads programmatically, with viewability not far behind. (Source: eMarketer)



With Incapsula reporting that 56% of web traffic is bot traffic (29% being bad bots such as impersonators, hacking tools, scrapers and spammers), these are legitimate concerns.



Bot fraud will cost the industry $6 billion in 2015

According to emarketer, global digital ad expenditures will jump 18 % in 2015 to reach $170.17 billion – 29.9% of the total advertising market. One third of it consists of bogus traffic – according to The Wall Street Journal.

  • Bots account for 11% of display ad views and 23% of video ads
  • Up to 50% of publisher traffic is bot activity, just fake clicks from automated computing programs.
  • Between 3% and 31% of programmatically bought ad impressions were found to be from bots, with an average of 17 percent.


“Fraud is a real and serious problem, but some, we think, still hold a mental image of fraudsters as one-off bad actors sitting in a dark room racking up clicks on ads on their site to make a few extra bucks. The truth is far more troubling: the majority of ad fraud today is perpetrated by sophisticated organizations that devote vast resources to build and operate large scale botnets run on hijacked devices, to reap multi-million dollar payouts” (Source: DoubleClick Publisher)

Do ads purchased programmatically even have the potential to be viewed by a human?

When ads are served that have not even have the potential to be viewed by a human user, it’s the advertiser who loses out. And this is a huge concern for the industry.

Advertisers want to buy viewable impressions – they want their ads to be seen! According to DoubleVerify, as reported on, RTB buying dipped from 48% viewable to 43% from Q3 to Q4 2014. The percentage fell from 55% to 51% on ad networks and from 64% to 58% on publisher-direct exchanges. So what can be done to increase the viewability of ads for advertisers? How can advertisers know they are purchasing impressions that have the potential to reach a human?

Digiday reported that Business Insider allows advertisers to purchase inventory based on viewability, but for a premium. But if advertisers have to pay more for a “viewable” impression, doesn’t it imply that the regular priced impressions are not viewable? A better solution for advertisers would be to take steps to increase viewabilty and eliminate fraud so that the concerns for all advertisers decrease, rather than simply charging more for “fraud-free” or “viewable” impressions.


The Wall Street Journal provides these tips to reduce the likelihood of getting your ads “seen” or clicked on by bots rather than humans:

  • Bots are nocturnal: Most human Web users tend to sleep during the night, but White Ops found that bot percentages spiked between 11 p.m. and 5 a.m, so advertisers might consider not running any ads during the night.
  • Bots like some types of content more than others: Sites related to finance, family and food had among the highest percentages of bot traffic, the study found, ranging from 16% to 22%. Tech, sports and science-related sites had among the lowest bot percentages, ranging from 3% to 4%.

Buzzcity recommends advertisers do the following:

  • Ask your suppliers what they are doing about ad fraud.
  • Use 3rd party technology to identify bots and ensure that you do not (re-) target them.
  • Go with your gut – if something looks too good to be true, then it probably is.

In case you need assistance, don’t hesitate to reach out to our team. We will gladly equip you with the right technical background to help you invest into human sources only.

One thought on “Digital advertising fraud concerns from media buyer’s aspect

  1. I do consider all of the ideas you’ve presented on your post.
    They are really convincing and can certainly work.
    Nonetheless, the posts are very short for beginners.
    May you please extend them a bit from next time? Thanks for the post.

Leave a Reply

Your email address will not be published. Required fields are marked *