We are now part of Schlesinger Group. Learn more
Request a quote
Request a quote
Blog

07 Feb 2020

How fraud is impacting the valuable data in market research

How fraud is impacting the valuable data in market research

February 7, 2020 / Events and blogs / Blog

Where It Hurts the Most

In a world where we find ourselves living online, it comes as no surprise that almost every aspect of business is executed digitally and in the cloud. The need for robust security systems is more critical than ever. While technology has driven innovation, it has also driven fraud which has impacted business on a global scale. Industry leading researchers have suggested that the Advertising arena alone could stand to lose upwards of $44 Billion by 2022.

The market research industry is no exception which is why it is so important to recognize the pain points. Fraudulent activity in the research arena is increasing as hostile players attempt to monetize the outcome of paid surveys. The most common fraudulent tactics are as follows:

  • Unengaged Respondents: Respondents blindly filling in surveys without reading
  • Deliberate User Fraud: For instance, responding to the same survey multiple times through multiple accounts by using different machines or IP addresses.
  • Bot Surveys: Surveys completed by applications designed to fake results
  • Vendor Manipulation: Survey information provided to respondent from a vendor to improve conversion, creating a too positive response.
  • False Completes: Faking completes in the vendor’s system without ever answering the client’s questionnaire.”

Behavior is Not Always What It Seems

While online fraudulent tactics are pervasive and must command attention, it’s also important to recognize that there can be false positives where the respondents suffer when our industry over corrects.  For example, if two people access a survey from the same IP, they will be marked as duplicates, even if they are unique individuals.  False positives are also created from overzealous models that seek to create the perfect data set. This happens when a survey taker prefers to take surveys at night, but the fraud prevention model has recognized real fraud often occurring at night. The legitimate survey taker may be marked as untrustworthy based on little evidence. Furthermore, the respondents who experience the false positives are burnt.  The messaging for sessions identified as fraud is left intentionally vague so as to dissuade real malfeasance. The survey industry can be taxing to respondents due to the inherent low conversion from click to complete, especially with additional hidden security rules.

Internet Bots and Organized Crime

Most researchers group all fraud into “bots”, but there are important differences.  Fraud can be broken down into 3 categories: DIY, bot, and clickfarms/advanced.

  1. DIY fraud:  This type of fraud is committed by individuals. They are respondents that engage in multiple surveys, entering false data that they feel will help them qualify. They research ways to improve their chance to earn a reward, even if it involves lying or light, technical hacks like URL manipulation. While DIY fraud is frustrating, it is not as frequent and is typically easy to identify and address.
  2. Internet Bots: Bots are the most recognized of all organized fraud. A developer or a small group of developers create a program that replicates respondents, in some cases a thousand plus. The program completes as many surveys as possible to earn the largest incentives, in an effort to cash out before they are caught. These bots are noticeable in survey data because the discrepancy and anomaly in their answers are obvious. Clear indications of this kind of fraud include:
    • Similar or the same time frame
    • Same answer pattern
    • Similar IP address
  3. While bots are dangerous, there are limitations to programs that allow advanced security systems to recognize them. Consequently, bots primarily prey on weaker security systems.

  4. Clickfarms: The worst type of fraud is organized, informed groups. As such, Clickfarms pose a real threat. Clickfarms are often located in countries where the cost of labor is exceptionally low, so there is an ROI on survey fraud. A group of real humans accessing information via hidden VPN’s can then share best practices and skirt basic security practices. Often times Clickfarm participants can evade CAPTCHA and similar tools that are developed to prevent bots. Unlike bots, humans also learn quickly, so a single security measure is not sufficient

Your Reputation is at Stake

Even small errors caused by fraud have an impact on the quality of the data. Large scale fraud can not only affect the integrity of the data, it can also impact the reputation of the research company. That said, the focus on fraudulent activity goes beyond the market research companies. Data Quality is important for all types of brand decisions in every aspect of business. Fraud affects the market research buyers equally as much as it impacts the research companies. While corporations use research data to influence the development of products, they also use data to provide insights into the behavior and preferences of their customers. It is critical that the data generated be dependable and actionable. Online trust in data collection is vital, and if fraudulent activity is pervasive, the integrity of both data and company will be breached.

Keep a Sharp Pulse

Clearly, as with any industry, fraudulent activity is evolving every day and because of the access to rewards which can often be monetary, the market research industry is particularly vulnerable. The consequences of a fraudulent attack results in poor quality data and biased research. What it comes down to is that market research companies face an extremely difficult challenge every single day. It is more important than ever to act with urgency and focus in an effort to maintain a great respondent experience while keeping a sharp pulse on the market to minimize and even mitigate ongoing fraudulent threats.