Categories
Artificial Intelligence and Machine Learning
March 22, 2021
Market research fraud is massive. fraud spiked in 2020, as much as 30% in some studies.
0
Fraud in market research continues to be massively problematic. Everyone has felt it. Exacerbated by the pandemic, our industry saw a big spike in fraud in 2020. On average across the industry, 15 to 30% of market research data is fraudulent. In some studies, the figures are even higher. We’re not alone; eCommerce, credit card, and identity fraud all went through the roof last year.
To preserve the integrity of our work, it’s important that we understand how market research fraud is evolving, that we commit to innovating new deterrents, and that we band together to fight it with an industry-wide response. In this spirit, let’s closely examine what is happening and what we can do about it.
First, let’s define the bad actors and what they are doing. Fraudsters can be defined as unqualified people who invade market research surveys to make sums of money from the incentives. In today’s world, they’re increasingly difficult to identify, and staying ahead of them requires vigilance.
We’ve long been plagued by bots, automated programs that drive large volumes of fake traffic to market research surveys, but now the human fraudsters are wreaking just as much havoc, probably more. We’re faced with a professional cottage industry where people are paid by criminal enterprises to perpetrate fraud. These “professionals” teach others how to qualify for surveys and to pass quality control checks. They provide step-by-step instructions in videos posted on YouTube, Facebook, and other social media. And they are not in just a few countries anymore — they’ve gone global.
Identity theft is also a major contributor to market fraud’s explosive growth. Thanks to a number of massive global data breaches over the past decade, creating fake identities has never been easier. Social security numbers are being bought on the black market. And mobile phone numbers can be bought for as little as 15 cents.
As a result, the common techniques our industry has used to validate panelists’ identities are no longer enough. This includes steps such as matching a name and address against a third-party source or using two-step phone authentication.
With the challenges we face, multiple layers of anti-fraud defenses are necessary. But how do we do this while protecting the quality experience we need to deliver to qualified, legitimate panelists?
For the good people, it is essential to offer a low-friction experience and to avoid treating them with suspicion. We know that engaging panelists with a good experience is directly correlated with data quality. The good news is there are ways to deliver a great experience while also keeping the bad actors out.
We enable identity validation at the individual panelist level by leveraging first-party panels. We use a balance of passive and active methods in multiple layers to understand who our panelists are. With passive checks like cross-checking identity with multiple third-party directories, the panelist can be totally unaware of your due diligence behind the scenes. Active checks like two-step phone authentication still have a place… but need to be combined with other things. At DISQO, we added “old-school” manual techniques a couple of years ago. Phone outreach by customer care teams for live verification and sending postcards with PINs to physical locations each have renewed relevance in the digital age.
Even with a combination of these techniques and many more, fraud can still get through. Bad actors will find ways to bypass almost any check. So we must continuously look to more sophisticated and innovative techniques.
It is promising that there is a burgeoning marketplace of commercial anti-fraud services. They each have their own “secret sauce” for detecting and mitigating fraud in real-time. They also bring the important advantage of being able to see and monitor the playing field across multiple sample providers within their platforms. These services are an important tool in the arsenal.
Artificial intelligence and machine learning offer much-needed automation to pour over large volumes of data — looking for suspect behaviors. We’ve enhanced our models to take into account highly accurate fraud flags that are actionable for preventative measures. We value that these are passive approaches with little to no impact on panelists’ experience.
This data science allows us to score the risk of an individual panelist and then apply the appropriate types of verification measures accordingly. This way, you can alleviate the need for good panelists to go through some of the more rigorous validation processes we would apply to others. Best of all, artificial intelligence models learn continuously and get smarter over time as they process more and more data.
The industry can also look to Fintech for powerful new solutions. After all, banks know a thing or two about identity verification. ID validation combined with facial recognition through biometrics is proving to catch fraudsters in their tracks, and it only takes three minutes for panelists to complete the process. This technique is highly effective for weeding out bad actors, viable internationally, CCPA and GDPR compliant, and can even be configured to accommodate people with disabilities.
Mitigating fraud is never-ending. Thieves will work just as hard to learn as we do, and they will always find a way — wherever there is marketing money to be siphoned online. This means the market research industry must commit to not only innovating new solutions but also to coming together to share best practices and learn from each other. We are a close industry, and I’m proud that we collaborate on the big issues. Let’s do more. It is essential we protect the value of market research to ensure our companies and clients can make business decisions with confidence.
Photo by Camylla Battani on Unsplash
Disclaimer
The views, opinions, data, and methodologies expressed above are those of the contributor(s) and do not necessarily reflect or represent the official policies, positions, or beliefs of Greenbook.
Comments
Comments are moderated to ensure respect towards the author and to prevent spam or self-promotion. Your comment may be edited, rejected, or approved based on these criteria. By commenting, you accept these terms and take responsibility for your contributions.
More from Bonnie Breslauer
Demand for insights about customer experience is more urgent than ever. Whether it’s in-depth strategic research led by insights professionals or cust...
The insights industry is facing a global supply shortage.
Is human recall reliable enough for conducting shopper research?
Top in Quantitative Research
Why are we still measuring brand loyalty? It isn’t something that naturally comes up with consumers, who rarely think about brand first, if at all. Ma...
Sign Up for
Updates
Get content that matters, written by top insights industry experts, delivered right to your inbox.
67k+ subscribers