Principal–Agent Theory and Affiliate Fraud: Why Performance-Based Marketing Creates Opportunities for Strategic Misreporting

In microeconomics, one of the critical concepts is the principal-agent problem, which refers to a situation in which an individual (the "principal") hires or contracts another individual (the "agent") to perform a task on the principal's behalf. Differences in information and incentives between principals and agents create moral hazards, whereby agents act out of self-interest while causing harm to the principal.

An excellent example of this is affiliate marketing. In a typical affiliate marketing relationship, an advertiser compensates its affiliates with a commission for either registering or making a purchase by clicking on the ad. The advertiser is the principal, and affiliates are the agents of the advertiser. Performance-based marketing creates a powerful incentive for strategic behavior by affiliates since affiliates are rewarded based on measurable outcomes rather than on directly observable effort.

The reason behind affiliate marketing fraud is also straightforward. The primary driving force behind affiliate fraud is information asymmetry. Affiliates know significantly more about how to create traffic, clicks, and conversions than the advertisers. As a result, advertisers cannot tell whether the conversion or the click was created by a real consumer or if the agent manipulated the attribution process to receive credit for that conversion.

From an economic perspective, affiliate fraud is an attempt to capture economic rents without creating the value that commission payments are intended to reward. Fraudulent affiliates exploit weaknesses in the tracking system and are therefore overpaid compared to what the affiliates contributed to the process. The payoff from committing this type of fraud can be significant, especially when it is difficult to detect fraudulent practices and enforcement costs are high.

One of the most famous cases of affiliate marketing fraud involved Shawn Hogan. Hogan was one of the highest-earning affiliates of eBay and reported earning over $28 million in affiliate commissions. According to court records and subsequent news reports, Hogan defrauded eBay by using a method known as cookie stuffing to place cookies on users' browsers without their consent. Whenever a user later purchased an item from eBay, Hogan received a commission, even though he had not built the relationship with that user. He was later prosecuted and sentenced to prison.

This case exemplified the agency problem perfectly. eBay intended to compensate affiliates for creating new customer relationships; instead, the affiliate exploited his information advantage over eBay and redirected the compensation to himself. What appeared to be outstanding performance was actually a transfer of value from the principal to the agent.

Another example of affiliate marketing fraud involves Fraud As a Service (FAAS). Some of the most notable cases have involved fraudsters creating thousands of bots and servers, enabling them to generate high levels of fake advertising activity and receive millions of dollars in payments for impressions and engagements from advertisers when in fact, no impressions or engagements had come from real customers

Affiliate fraud is one manifestation of a broader problem of information asymmetry in digital markets. In addition to the costs incurred by advertisers who become victims of affiliate fraud, advertisers also incur significant costs when recruiting, screening, and managing new affiliates. From the perspective of principal–agent theory, investments in technologies that detect affiliate fraud represent a form of monitoring expenditure designed to reduce information asymmetries between advertisers and affiliates and limit opportunities for strategic misreporting.

As advertisers attempt to limit fraud, they often impose stricter rules and lower commission rates, which can reduce earnings for legitimate affiliates. Additionally, the emergence of affiliate fraud has been a factor in reducing the efficiency of the overall digital advertising marketplace.

This issue has a significant scope. Digital advertising fraud is estimated to cost advertisers between $20 billion and $30 billion annually according to research done by a number of industry and marketplace analysts. Many are now predicting that losses resulting from global digital advertising fraud will exceed $170 billion per year by the end of this decade. Although the majority of costs, losses, and damages resulting from the fraud will occur through digital advertising channels that originate outside of the affiliate program, an increase in the occurrence of affiliate fraud is indicative of a larger trend: the difficulty of accurately measuring and compensating for performance in the online marketplace.

The reason affiliate fraud continues to persist is that it represents a direct product of the principal-agent problem described above; when an agent’s incentive compensation is tied to measurable performance, there exists an incentive for an agent to manipulate the system to increase their expected performance. As the economy becomes increasingly data-driven, advertisers must ensure that performance metrics accurately reflect value creation rather than opportunities for agents to increase compensation through manipulation.