Leveraging advanced analytical techniques to combat ad fraud can lead to significant improvements in advertising efficiency, benefiting various industries.
Kumar
In a rapidly evolving digital landscape, understanding and implementing effective strategies to combat ad fraud is essential. There are several practical applications of experimental techniques in ad fraud detection, with a focus on evaluating the success of reducing ad spend as a preventative measure.
ADVERTISEMENT
Leveraging advanced analytical techniques to combat ad fraud can lead to significant improvements in advertising efficiency, benefiting various industries. Collaborative efforts between data scientists and marketing professionals can result in valuable knowledge exchange, empowering teams to uncover and address fraudulent activities in advertising. For example, senior analytics leader Saurabh Kumar has successfully merged data science expertise with real-world marketing applications. As a prominent figure in AI driven marketing analytics in consumer tech, Kumar has made significant strides in identifying ad fraud through experimental approaches, such as reducing ad spending to evaluate its effects. In a detailed discussion, Kumar shared insights on how these methods can help organizations optimize their marketing budgets and enhance the integrity of digital advertising practices.
While working for Uber, Kumar designed experiments to assess the impact of reducing the marketing budget on new rider activations, which led to the discovery of over 100 million USD in non-productive ad spend, significantly boosting the company's savings. His findings, which have been cited in numerous industry publications, marked the first major exposure of deceptive practices such as ads on forbidden sites and fabricated reports. Under Kumar's guidance, Uber recovered over 6 million USD in damages from the 2019 lawsuit, highlighting the severity of ad fraud in the industry. This settlement not only validated Uber's claims but also set a precedent in the ad tech sector, emphasizing that not all advertising expenditures yield positive returns. Kumar's work has been instrumental in raising awareness about the widespread nature of ad fraud, encouraging advertisers to scrutinize their ad investments more closely. The allegations included fraudulent activities such as displaying ads on prohibited sites, employing deceptive practices like auto-redirects, and fabricating ad placement reports.
Kumar recounted an instance where Uber faced escalating ad costs due to competitors increasing their bids on Uber's branded keywords in the ridesharing market. The company's response was to similarly increase its ad spend on these keywords, potentially exceeding $24 million in marketing expenses for acquiring new drivers in North America. To assess the effectiveness of this strategy, Kumar and his team implemented a matched market test across several clusters of cities. The results revealed that the increased spending on branded keywords did not lead to a corresponding increase in driver activations. While the additional spend did boost vanity metrics like initial sign-ups, it failed to generate meaningful long-term conversions. This critical insight prompted the team to pivot their strategy, redirecting the budget to more effective channels, ultimately saving Uber over $24 million annually.
Kumar concluded by stressing the importance of professionals sharing their experiences with academia and industry peers through research publications and conference presentations. Such knowledge exchange fosters the discovery of new insights and facilitates the industry's advancement. By learning from one another's successes and failures, these collaborative efforts can drive meaningful progress and innovation.