Our Fight Against Market Research Fraud: Using Next-Gen Fraud Defenses | Marketing Maestros | Blogs | ANA

Our Fight Against Market Research Fraud: Using Next-Gen Fraud Defenses

July 22, 2021

By Bonnie Breslauer

Sorbetto/Getty Images

There is an immense focus on cookieless solutions as third-party cookies begin to crumble. Given the changing methodology, the industry must continue to be cognizant of fraud and the impact it has on media effectiveness research. As any experienced market research professional knows, fraud is something that the general market research industry has had to combat for years, if not decades.

The ANA is renowned for its bold stance on digital advertising fraud — effectively the “last mile” in reaching the consumer with commercial messages. But what if your digital buying tactics are based on fraudulent data? This affects planning for frequency, reach, creative/placement rotation or sequencing — you get the idea. Critical decisions and significant ad spend can be wasted, or even worse, have a negative impact on the brand.

Because market research fraud is an industry-wide problem, we believe industry leaders should collaborate around solutions when possible. In industry forums, we’ve been championing collaboration to protect our industry’s unique value and ensure brands can make sound decisions.


Fraudsters Are Getting Bolder

Fraudsters are getting better at what they do, something the industry realized the hard way when market research fraud spiked during the pandemic. Imperium, an anti-fraud service provider, reported a 200 percent increase in fraud last year.

DISQO’s audience insights platform and many consumer panels invite participants to share opinions in exchange for compensation. The allure of a cash reward can be catnip for fraudsters. We were not surprised when the industry was targeted during the tough times of 2020, especially given the increasing sophistication of the thieves who’ve gotten quite savvy in working around industry defenses. And if one loophole is closed, they’ll find another.


Combining Passive and Active Fraud Detection Can Protect Panelist Experience

Data quality is of paramount importance. Whether you’re measuring brand lift or conducting market research, the threat is existential.

The good news is that there are many tools in the researcher’s kit to identify fraud. These include anti-fraud services that detect potential fraud in real-time, typically using device fingerprinting to look at properties associated with the device. The use of in-survey checks is considered a best practice. These consist of “trick” questions which are inserted into screeners and surveys. Patterns such as speeding or straight lining responses are monitored as well.

Market researchers also identify fraud through nuances in the data. For example, you can spot nonsensical responses to open-ended questions or data that doesn’t align with other benchmarks. For longitudinal research, inconsistent data trends may also signify a concerning issue. The rub here is that data issues are found late in the game — many times after fielding has closed. The “scrubbing” process can be time consuming for the researcher and can cost brands precious time in their decision-making.

At DISQO, we’ve enhanced our machine learning models to assess panelists for the likelihood of being fraudulent, essentially giving them a score based on their behaviors within our site. We’re also implementing innovative fraud detection practices that, when combined with extensive verification checks, create an effective mix of anti-fraud deterrents.

For example, we leverage our first-party panel to deploy passive techniques conducted without user involvement, such as data verification through multiple third-party databases, plus active measures like two-step phone authentication. We even mail postcards with pin codes that panelists enter to verify their identity and physical location. In our digital world, analog methods can still be quite effective.

This is quite a balancing act. One of our biggest challenges is to create an engaging and rewarding experience for our panelists while still weeding out the bad actors. Our verification practices consider this dichotomy as they are applied.


Adding Layers of Technology Innovations

DISQO has also started using a new Fintech validation technique relied on by banks that we’re excited to introduce to the market research industry. It’s proven to be successful in the financial arena, and we think it has great potential in our industry as well. We’re leveraging ID validation and facial recognition through biometrics. It only takes about three minutes for a good panelist to complete the verification process, and it’s highly effective at weeding out bad actors.

Because of the rise in the volume of fraud and the increased sophistication we’ve seen, we believe the industry will need to continually evolve technology and processes to stay a step ahead. It is not easy, but it is also not optional given the crucial nature of sound insights for critical business decisions. The thieves won’t stop innovating, so neither can we.

Bonnie Breslauer is chief customer officer at DISQO.

The views and opinions expressed in Marketing Maestros are solely those of the contributor and do not necessarily reflect the official position of the ANA or imply endorsement from the ANA.

You must be logged in to submit a comment.