Billion-Dollar Decisions at Risk: The Alarming Rise of Survey Fraud

Every year, brands spend billions of dollars on market research, aiming to understand consumer sentiment about issues at large or to know specific customer perspectives on their products. It can help those companies make critical decisions about product development or placement and influence how they market their products through advertising. 

From that industry, we estimate that $1 trillion worth of business and advertising decisions are made based on information gleaned from surveys. Yet, many of those surveys contain too many duplicate users and bots, driving an unacceptable percentage of bad data due to fraud or poor quality. 

In a recent study of over 200,000 recent survey sessions bought from the top three panel providers, we’ve found over 40% of surveys are bad quality. This seriously calls into question any important decision made as a result.  

How Survey Fraud Occurs  
Surveys only work and produce meaningful results if each survey response is a unique human being who pays attention and fills out the survey accurately. Where fraud occurs is if people attempt to fill out the survey multiple times, therefore falsely inflating their answers, or someone uses technology to trick the system into thinking it’s a human filling out the form. 

Financial Incentives  
Panels often pay “users” to take surveys as it's seen as a worthwhile value exchange for customer perspectives and is the only way to generate scale. There are plenty of people who follow the rules by only filling out a survey once and actually paying attention to questions and answering honestly.  

But there is an unacceptable number of people who try to game the system by taking surveys multiple times, thwarting any simple quality control by pretending to be different people. Of course, this is a laborious process that provides the fraudsters more money, but not enough to make it the industry’s biggest problem. 

Bad Actors and the Rise of Technology to Enhance Fraud  
More concerning is the fraud farms that use technology like generative AI-powered bots and other automation tools to fill out multiple surveys quickly. Doing so produces significant revenue and takes little human effort once the technology is set up. 

These bad actors can use that technology to create hundreds or thousands of survey responses in seconds, in ways that will pass the panels’ quality checks or specific criteria.  

Now that we’ve determined that survey research is riddled with fraud and quality issues, in our next post, we’ll discuss why survey providers are falling behind in tackling this issue.