How Data is Aiding the Fight Against Ad Fraud
Concerns over advertising fraud are on the rise, and with good reason. The amount of lost revenue on fraudulent web traffic is expected to soar to $16.4 billion, according to research by WPP. That number is more than double the $7.2 billion the Association of National Advertisers estimated would be lost due to online ad fraud just last year. While it’s often difficult for advertisers and publishers to quantify the financial impact fraud has had on their business, the industry has started working together to combat the issue.
Types of Ad Fraud and the Bots that Cause Them
The days of the shady fraudster operating out of a desolate basement clicking on ads is long gone. Believe it or not, ad fraud is well on its way to becoming one of the largest organized crime businesses in the world, second only to the drug trade. “It's really the hi-tech, well-funded criminal networks that are presenting the greatest challenge to the digital advertising ecosystem,” said Mike Zaneis, President and CEO of the Trustworthy Accountability Group (TAG).
So, why on earth can’t this be stopped? Well, defeating ad fraud involves more than just shutting down a few fake sites. Even legitimate sites (with huge ad exchanges and fraud-busting teams) get caught in snares set by online criminals. These scoundrels are sophisticated. When they cook up fraud, they do it at the code level – something beyond your garden-variety marketer’s tinkering ability.
Just how sneaky does ad fraud get? Expert Dr. Augustine Fou identified two main types: impression fraud and search click fraud. At the heart of both varieties are “bots,” or software applications that run automated tasks over the Internet.
- Impression fraud occurs when bots are used to repeatedly load pages, mimicking human browsing behavior and generating fake ad impressions.
- Search click fraud happens when bots type keywords in search boxes, causing search ads to load; then, these bots click on the ads, generating cost-per-click (CPC) revenue.
Soft conversions (via clicks, views, etc.) are at higher risk for ad fraud because it’s much easier for bots to replicate this activity than, say, filling out a registration form.
Online criminals like programmatic placements because they’re automated and less susceptible to scrutiny. And if these programmatic purchases are for rich media (such as video), so much the better, as they carry a higher cost per thousand (CPM) and promise a bigger pay day to fraudsters.
“Criminals are always looking for the path of least resistance, and unfortunately, they found our industry,” said Zaneis. “They've really made a home here because it's been highly profitable. So, there are very sophisticated, well-funded, organized criminal networks that are perpetrating the largest, most complex ad fraud schemes.”
Ad Fraud Has Evolved
We may like to believe that this illicit activity happens far away from us, originating from criminals’ computers in a remote country. But actually, most bots come from residential IP addresses, many of which lead to unaware consumers. Often, bots are inadvertently installed when users click on a seemingly legitimate ad, download a toolbar, or install a browser extension. Once a computer is infected, bots run silently in the background, acting uncannily human by visiting multiple sites and collecting cookies.
You may recall the large-scale Methbot attack, a sophisticated take on traditional click fraud that wreaked havoc on the ad industry. “This attack was different in nature and scale than previous scams, and one of the largest the industry had ever seen,” said Zaneis. Some reports estimate that the criminals were netting between $3 and $5 million a day by siphoning legitimate advertising revenues from over 6,000 websites. Fortunately, TAG played a major role in putting a stop to the Methbot operation.
“Working together with White Ops, we activated the information-sharing infrastructure we built over the last two years and organized an emergency briefing on the fraud operation for more than 180 anti-fraud executives at leading companies across our industry, so they could evaluate and respond to this threat in near real time,” explained Zaneis. “We also expedited our review of the IP addresses involved for inclusion on TAG’s blacklist of data center IP addresses that are significant sources of fraud.”
Thanks to TAG and actions like this, the fraudsters are on guard. Founded in 2015, TAG is a cross-industry group born out of the 4A’s, ANA, and IAB to spur transformational improvement across the digital advertising ecosystem, focusing on four core areas: eliminating fraudulent digital advertising traffic, combating malware, fighting ad-supported Internet piracy to promote brand integrity, and promoting brand safety through greater transparency. But even with organizations like TAG, the industry has a big fight ahead of itself.
Who is Impacted by Advertising Fraud
Considering that digital scams siphon half of the dollars out of the advertising ecosystem, it shouldn’t be a surprise that relationships get seriously damaged along the way.
An overview of the fallout – and each party’s grievances – goes something like this…
- Buyers, already annoyed by irrelevant and invasive ads, become haplessly complicit in digital ad fraud; they fight back with ad blockers.
- Advertisers wonder how many actual human beings actually see their ads; they grow suspicious of programmatic placements in general (and media agencies that buy them).
- Digital publishers (even premium ones) lose credibility with skeptical advertisers and, in the process, lose valuable revenue.
Ad Fraud Detection and How to Protect Yourself
The first thing B2B marketers can do to combat ad fraud is to establish clear and open communication with their media-buying agencies. Don’t play the blame game; instead follow through on your due diligence. The ANA recommends asking tough questions, like:
- What third-party traffic validation tools do you use? Empower your agency to deploy bot and domain detection technologies to track traffic sources.
- Do you require publishers to identify all third-party traffic sources? Your agency can require that this language be built into requests for proposals (RFPs) and insertion orders. You can also reject sourced traffic and run your ads on organic site traffic instead.
- Do you use blacklisting? If so, how often do you update your list blocked traffic sources?
- What type of data do you prefer to use for B2B targeting? Probabilistic, deterministic or a mix of both?
Is Blockchain The Answer to Ad Fraud?
One recently proposed ad fraud/safety solution that’s received a lot of media has been blockchain. Often mentioned in conversations about cryptocurrencies like Bitcoin, blockchain is essentially a ledger of records linked cryptographically and shared across a network of computers. Blockchain technology records transactions and “seals” them with a key code that makes it almost impossible to manipulate. This level of security coupled with blockchain’s ability to record comprehensive details on every step of an advertising campaign is appealing to marketers. But is it ready for prime time? Not so, say B2B marketers.
Only 5 percent said current blockchain solutions show promise, according to a report by Adweek and Dun & Bradstreet. But over 40 percent said they just don’t know. It looks like the situation where a buzzword has taken hold, but the market still doesn’t know if it’s the right answer.
One thing that is helping marketers in their fight against ad fraud is data.
Data’s Role in Combating Ad Fraud
The types of data a media-buying agency uses to identify potential customers directly impacts the effectiveness of the digital advertising. Probabilistic data uses anonymous data points to identify targetable prospects. These pieces of data are fed into a proxy model, which then spits out a percent probability that the user matches a desirable buyer profile. Deterministic data is gathered from verified sources and vetted for quality. It isn’t derived from assumptions. It’s collected from real business professionals – aggregated, edited, and verified.
Deterministic data offers a clear advantage for marketers, because it helps to confirm that prospects are actual human beings (and not bots). Based on this benefit alone, it’s easy to see why 32% of marketers planned to use deterministic targeting in 2016.
However, the value of deterministic data goes beyond verifying buyers. It’s also helpful for identifying legitimate (and fraudulent) businesses throughout the digital advertising ecosystem.
The Price Tag on Trust
Great strides have been made to foster trust throughout the online advertising supply chain, with TAG leading the way. (Disclosure: Dun & Bradstreet is a TAG partner.) As part of its charter, TAG qualifies digital advertising industry participants through a proprietary background check process powered by Dun & Bradstreet. After companies are confirmed as credible digital ad industry participants, they’re listed in the TAG Registry. Registrants also receive a TAG-ID, a unique global identifier certifying that their business is legitimate. Using TAG-IDs as a guide, B2B marketers can choose to work with only credible businesses, reducing the money flow to fraudsters in the process.
“At the end of the day, no one company can solve this alone, including us,” said Zaneis. “So, we help drive a lot of collaboration around information sharing about threats and developments of new technologies and protocols and help the industry collectively fight fraud and keep the money from going to the criminals. It’s really a crowd-sourced effort. We have something called our data center IP list and we source information from 12 to 15 major ad platforms every month about known sources of non-human traffic. We launched that a little over a year ago with about 500,000 IP addresses. These are IP addresses that only generate non-human fraudulent traffic. Today that list has over 65 million IP addresses and that only happens when you have a robust information sharing system.”
It’s this level of collaboration and transparency that has been crucial in helping thwart ad fraud. “We work closely with our members and other industry watchdog groups to openly share this data and intelligence,” explained Zaneis. “Given the advanced features of many of these ad schemes, we believe TAG's information sharing platform allows responsible industry actors to mitigate potential threats quickly and effectively.”
Data like this is helping police the ad industry. In addition to keeping track of IP addresses, there are other data points and algorithms the industry can use to identify deviant activity and abnormal trends. “Things like spikes in activity at certain times in the day when you wouldn't expect it, or from specific IP addresses or different regions of the world when you suddenly see a spike in activity from, say Southeast Asia, when your website doesn't do a lot of business there; there are a lot of red flags and data helps you identify them,” said Zaneis. “Again, it's really about sharing those data points and red flags across the industry.”
Relying on so many companies to help tackle the ad fraud problem means TAG needs to understand and verify the buyers, sellers and intermediary companies applying to participate in their registry. “When you’re jointly working to tackle something as big as ad fraud, the first thing you must do is create transparency across the ecosystem,” stated Zaneis. “Again, data plays a big role there.”
“We partnered with Dun & Bradstreet to essentially create the ‘White Pages’ for the digital ad ecosystem,” explained Zaneis. “D&B powers our background checks to ensure companies registering with us are legitimate and that we have an accurate idea of who the right contact within the organization is that can act as TAG’s compliance officer,” stated Zaneis.
This level of detail and accuracy is made possible by D&B’s proprietary business identifier known as the D-U-N-S® Number, which helps TAG keep the focus on fighting fraud without worrying any of its members are not legitimate. “Because the supply chain is so global in nature, we also require an international database of businesses and business contacts to fight ad fraud on all fronts,” said Zaneis. “D&B has been a been a key partner in our ability to really grow our registry across the globe.”
Marketers Are in This Together
It’s important to remember that fraud is not a publisher problem or advertiser problem, it’s an industry problem. Only by working together and remaining vigilant can the industry triumph over the bad guys. Zaneis was sure to reiterate this several times. “Everybody has a responsibility in getting us to a more trusted ecosystem. We’re not a fraud vendor, our mission is to bring the industry together to share data and information that helps us operate in the safest environment. Companies that can achieve this receive our verified fraud seal, and we hope to see every company earning this as we move forward. There's a synergistic and network effect of having more people fight fraud. So, growth is key for us.”