Interviews, insight & analysis on digital media & marketing

Outsmarting ad fraud: what every marketer needs to know

By Aurelie Guerrieri, Chief Marketing and Alliances Officer, DataDome

In my two-decade tenure as a marketer, I’ve managed ad budgets ranging from the modest to the mammoth, at times overseeing hundreds of millions in ad spend. Through this journey, one consistent adversary has been ad fraud. The allure of online revenue has birthed an ever-evolving battleground of deception, where ad fraud has become an unwelcome yet persistent participant.

Despite assurances from major ad networks and DSPs claiming to have ad fraud under control, the reality often tells a different story. I’ve witnessed first-hand businesses hemorrhaging tens of thousands in ad spend to sophisticated fraud schemes that slip through the cracks of conventional detection methods.

Ad fraud has morphed significantly over the years, evolving from simple impression and click fraud to more elaborate schemes involving app downloads and lead generation fraud. The stakes are high, with losses not just in money, but also in customer trust, decision-making quality, and brand reputation.

Today’s fraudsters wield AI with precision, deploying sophisticated bots that mimic human behavior with alarming accuracy. These bots, often operating from residential IPs, can easily bypass traditional detection systems, making them a formidable foe for any marketing team.

So what’s a marketer to do?

A Stark Reality

A delay in detecting and blocking ad fraud can result in considerable financial ramifications. In a matter of minutes, fraudulent clicks and engagements can accumulate, draining marketing budgets without providing any genuine value. According to Juniper Research, 22% of all online advertising spend was lost due to ad fraud in 2023, and the total amount of money lost is projected to exceed $170 billion in five years. 

On one occasion, we uncovered a significant volume of ad fraud in a customer’s campaign that had previously gone undetected, even though they were utilizing well-known ad networks. They were shocked to find that up to 35% of their ad traffic was bot-driven fraud that would never convert. Not only did this bot traffic drain their marketing budget, it resulted in poor campaign performance and caused time-consuming “clawbacks” with ad networks. 

This revelation underscores the necessity of leveraging AI to counteract fraud in real-time, emphasizing the need for an independent party to scrutinize ad traffic thoroughly.

Afterall, as marketers, we leverage artificial AI to enhance efficiency and streamline operations, ensuring our strategies are both innovative and impactful. However, the same advancements in AI that empower us are also being utilized by fraudsters, significantly increasing the complexity of combating ad fraud. This evolving challenge underscores the necessity for sophisticated solutions that can adapt and respond to the tactics employed by these malicious actors. 

Guiding Principles for CMOs

As a CMO, it’s important to maintain a broad perspective to effectively manage the complexities of ad fraud. This approach helps in identifying overarching patterns and strategies, ensuring we can address challenges comprehensively and safeguard advertising efforts. 

  • Challenge the Status Quo: Question the vested interests of your current ad fraud tools. Are they truly independent, or are they too closely tied to the platforms you’re advertising on? 
  • Deep Dive into Data: Analyze trends over extended periods and across vast datasets to identify fraud that has outsmarted your systems; given that fraudsters are adept at trickling in bad traffic in a good traffic stream, it takes a very large dataset to identify suspicious behaviors. Look for anomalies and patterns that don’t align with expected user behavior.
  • Align Detection with Fraud Incentives: Tailor your fraud detection strategies to your specific business incentives. For instance, if it becomes known that you blacklist publishers sending over 20% fraud, ensure mechanisms are in place to detect any attempts by these publishers to re-enter under a new guise or blend fraudulent traffic to just under your threshold.
  • Understand the Fraudster’s Playbook: For example, imagine your campaign is centered around mobile app installs with rewards tied to in-app activities. It’s critical to understand how bots might mimic human behavior to disrupt the campaign. Grasping the goals of fraudsters allows you to more effectively protect your campaigns from their strategies.
  • Leverage the Power of AI: An AI-powered online fraud & bot management solution provides a swift and robust answer to the ad fraud problem. The right solution will leverage multiple layers of machine learning techniques, including signature-based detection, behavioral analysis, time series analysis, and more, to distinguish automated traffic from genuine user interactions. This will empower advertisers to identify – and block – fraud faster and more accurately.

As CMOs, our goal is not just to report ad fraud but to outsmart it. By leveraging advanced, AI-powered solutions and adopting a strategic, data-driven approach, we can protect our marketing investments and ensure that every dollar spent contributes to genuine engagement and revenue growth. With the right solution in place, we can say goodbye to the days of manually sifting through spreadsheets to reconcile ad spend, and engaging in tedious negotiations with publishers for refunds on fraudulent clicks.