Jeromy Sonne, a seasoned veteran of the digital advertising landscape, has transitioned from mastering the intricacies of Facebook ads to becoming a leading authority in ad fraud detection. His firm, Daypart, now serves as a crucial advisory partner for advertising agencies and major advertisers, meticulously scrutinizing the accuracy and legitimacy of their campaigns. In a recent, in-depth conversation, Sonne peeled back the layers of the pervasive ad fraud problem, offering invaluable insights into its multifaceted nature and providing advertisers with actionable strategies for prevention and detection.
The Evolving Threat of Ad Fraud
Ad fraud, Sonne explains, is not a monolithic issue but rather an expansive umbrella term encompassing a range of deceptive practices. At its most basic level, it involves the deliberate billing of bot-generated impressions or automated clicks, masquerading as genuine engagement. However, the sophistication of these schemes extends far beyond simple automation. Sonne highlights instances where advertisers pay premium rates, such as a $30 CPM (cost per mille, or cost per thousand impressions), for ad placements that are ultimately served on low-quality, irrelevant websites, thereby devaluing the intended reach and impact of the campaign.
"The spectrum of ad fraud is incredibly broad," Sonne stated. "From simple bot traffic designed to inflate impression counts, to more elaborate scams that trick advertisers into believing their ads are reaching desirable audiences when in reality they are on obscure, low-tier sites. And at the most extreme end, we’re seeing ad fraud being used as a tool for money laundering, where criminals essentially create their own shell companies to buy ad space from themselves, legitimizing illicit funds through seemingly legitimate transactions."
This latter point, the use of ad fraud for financial crimes, underscores a growing concern within the industry. The sheer volume of programmatic advertising, which facilitates automated buying and selling of ad space in real-time, creates fertile ground for sophisticated criminal enterprises to operate. While platforms like Meta and Google have implemented robust internal controls, the more open and fragmented nature of programmatic networks presents a more challenging environment for fraud detection and prevention.
Programmatic Networks: The "Wild West" of Ad Fraud
When questioned about the prevalence of ad fraud across different advertising platforms, Sonne drew a clear distinction. "While no platform is entirely immune, the problem is significantly more acute within the realm of programmatic advertising," he explained. "These are the open exchanges where ads are bought and sold at scale, often across a vast and less regulated network of websites and apps. In contrast, the ‘walled gardens’ like Meta and Google, while not perfect, tend to have more controlled environments and are generally more responsive to fraud concerns."
Sonne elaborated on the unique challenges posed by programmatic networks, often referred to as the "Wild West" of digital advertising. "The complexity of the ecosystem, with its numerous intermediaries and automated processes, makes it easier for bad actors to hide their activities," he noted. "Crooks can operate sophisticated scams, often evolving their tactics to evade detection. We might see an individual posing as a legitimate advertiser, using flashy imagery of luxury items like Lamborghinis to attract initial attention, but the underlying operation is fraudulent. Or, a seemingly legitimate offer can gradually morph into a scam over time."
The implication for advertisers is stark: cutting out fraud is one of the most effective ways to boost return on ad spend (ROAS). "The most fundamental optimization tactic available to advertisers is simply ensuring that their ads are reaching real humans, not bots or fraudulent entities," Sonne emphasized. "When you eliminate wasted spend on illegitimate impressions and clicks, the real value of your legitimate ad placements naturally increases."
Unmasking Deception: Strategies for Detection
For advertisers like Eric Bandholz of Beardbrand, whose company dedicates its entire ad budget to Meta, understanding the potential for fraud within even seemingly secure platforms is critical. Bandholz posed a direct question: "What percentage of impressions or clicks are from bots or unscrupulous pages?"
Sonne acknowledged that precise figures can vary significantly depending on the specific channels and networks used. "On Meta, for instance, you’d want to pay close attention to its Audience Network," he advised. "This is where ads are extended to mobile apps and other third-party placements, and it can indeed be a higher-risk environment. Meta does proactively identify and remove fraudulent activity; they might purge a significant percentage of clicks or impressions that are deemed to be from bots or automated agents. This can include AI-driven bots or even compromised human accounts being used for fraudulent purposes."
However, Sonne reiterated that the open web, encompassing display ads, video ads, and mobile apps outside of the major walled gardens, remains a far greater concern. "Estimates vary, but it’s not uncommon for upwards of one-third, or even more, of ad spend on programmatic networks to be subject to some degree of fraud," he stated. This figure, if accurate, represents a staggering loss of potential return for businesses investing in digital advertising.
The question then becomes: how can advertisers discern fraudulent activity from legitimate campaign performance? Sonne’s answer centers on vigilance and a commitment to transparency. "Transparency is the absolute key," he asserted. "With sufficient data, you can begin to identify anomalies and inconsistencies that signal fraudulent activity. It’s about looking beyond the surface-level metrics provided by ad platforms and digging deeper."
He provided a compelling anecdote from his own work: "I was recently reviewing a set of mobile ad campaigns that, at first glance, appeared to be performing adequately. However, when I investigated the devices on which these ads were being served, I found a significant number were reported as iPads with Intel chips. This is a red flag, as iPads have never been manufactured with Intel processors. This clearly indicated publisher spoofing, a workaround to deceive tracking mechanisms, or potentially the operation of bot farms designed to mimic legitimate device usage."
This "Sherlock Holmes" approach, as Sonne described it, involves actively seeking out patterns, claims, or data points that simply don’t add up. Even within platforms like Meta, advertisers can take proactive steps. "You can implement measures like blocking specific shady apps on the Audience Network, or even deploying your own ad server to cross-reference data with what the ad platform reports," he suggested. "Any discrepancies should be flagged and investigated. There’s no magic bullet, but the principle of ‘trust but verify’ is paramount. When you identify a discrepancy, challenge it with the ad platform and work towards obtaining refunds for fraudulent activity."
When to Call in the Experts: The Role of Ad Fraud Detection Firms
The complexity and resource-intensive nature of ad fraud detection naturally lead to the question of when it becomes prudent to engage specialized firms like Daypart. Sonne outlined the typical client profile for his services.
"Generally, I work with agencies that are managing significant advertising budgets on open web programmatic ads, including connected TV and audio," he explained. "This is typically for clients who have already optimized their spend on Meta and Google and are now venturing into these more complex, higher-risk channels. Companies investing tens of millions of dollars annually can justify the cost of our specialized services."
For advertisers whose entire budget is confined to Meta and Google, Sonne indicated that the threshold for engaging external fraud detection services might be lower, given the relative safety of these walled gardens. "These platforms police their own ecosystems more effectively than others," he acknowledged. "I have a great deal of respect for companies like Comcast, and their programmatic buying platform, Beeswax, is known for its tight controls and rigorous operations."
However, even within these more secure environments, the potential for nuanced fraud exists. The key takeaway is that as advertising spend scales and diversifies into more complex channels, the need for expert oversight in fraud detection becomes increasingly critical. The goal is not just to identify and recoup losses, but to build more robust, efficient, and ultimately more profitable advertising strategies.
The Future of Ad Fraud and Advertiser Responsibility
The conversation with Jeromy Sonne offers a sobering yet empowering perspective on ad fraud. It highlights that while the problem is pervasive and sophisticated, it is not insurmountable. The rise of AI and machine learning in advertising presents both new avenues for fraudulent activity and powerful new tools for its detection.
The increasing reliance on data analytics and the development of more sophisticated tracking and verification technologies are crucial. However, Sonne’s emphasis on fundamental principles like transparency, vigilance, and a healthy skepticism remains the bedrock of any effective fraud prevention strategy.
As the digital advertising ecosystem continues to evolve at a breakneck pace, advertisers must remain adaptable and informed. The insights provided by experts like Jeromy Sonne are invaluable in navigating this complex terrain, ensuring that marketing investments yield genuine results rather than enriching fraudulent actors. The ongoing battle against ad fraud is a testament to the dynamic nature of the digital world, where innovation and deception constantly vie for dominance, and where informed diligence is the advertiser’s most potent weapon.
