By Frank Kelly and John Bremer. Frank is Market Research Practice Lead at Virtual Incentives and John is Advisor at Tenetic. They are members of the IA Standards Committee.
Survey fraud, defined as a systematic act of intentional deception in order to gain a benefit during the conduct of survey research, has moved from being an occasional nuisance to an existential challenge for the market research industry.
Industry efforts such as the Global Data Quality (GDQ) Initiative have highlighted the scale of the challenge, with studies consistently showing that a meaningful share of incoming survey traffic exhibits fraudulent or low-quality characteristics. In many cases, researchers report removing double-digit percentages of completes during data cleaning—underscoring both the prevalence and growing sophistication of the issue.
Against this backdrop, the Insights Association Code of Standards and Ethics provides a clear mandate: researchers must ensure transparency, methodological clarity, and data integrity. Nowhere is that more relevant than in how we prevent and disclose acts of fraud.
Bringing It Back to the Code
Fraud prevention is not just a technical function—it is a direct expression of professional standards.
By expanding fraud detection across the full research lifecycle—and being transparent about it—we reinforce core principles of the Insights Association Code of Standards and Ethics:
- Duty of Care (Section 1): Taking reasonable steps to identify and mitigate risks to data quality, including fraudulent participation.
- Primary Data Collection (Section 2): Applying appropriate and well-considered methods—and clearly communicating them—to ensure the validity of collected data.
- Integrity (Section 4): Safeguarding research results so they reflect real, qualified participants rather than fraudulent or duplicated responses.
- Reporting (Section 7): Providing clients with sufficient transparency about data quality controls to support proper interpretation of findings.
Why Transparency Matters
Fraud prevention has often been treated as a “black box” handled by sample providers or technology partners, but today’s environment demands more openness.
The Code is explicit in regard to how organizations should handle survey fraud in their ecosystems.
● Section 2 – Transparency requires researchers to “provide clear, accurate, and appropriate information on research methods and findings.”
● Section 4 – Integrity requires that researchers ensure data quality and not misrepresent findings.
If fraud detection is invisible, stakeholders cannot properly assess the reliability of the data. Transparency is not just good practice—it is essential to ethical research. Organizations are responsible for the actual detection and elimination of fraud in their survey execution, to the best of their abilities, but also for allowing users of the information to understand how they have done so.
What Transparent Fraud Prevention Looks Like Today
Across the ecosystem there is growing convergence around several core practices.
1. Multi-Layered Detection, Not Single Checks
Fraud prevention now relies on layered methodologies:
● Device fingerprinting
● IP and proxy/VPN analysis
● Behavioral analytics
This aligns with Section 2 – Research Design, which requires the use of appropriate and robust methods.
2. Behavioral and Network Intelligence
Modern fraud detection extends beyond a single survey:
● Monitoring participation patterns across studies
● Identifying coordinated activity
● Detecting masked or synthetic identities
These approaches support Section 2 – Primary Data Collection, by applying appropriate methods to detect and prevent invalid participation, and reflect researchers’ Duty of Care (Section 1) to take reasonable steps to protect data quality.
3. Pre-, In-, and Post-Survey Controls
Leading practices now span the full research lifecycle:
● Pre-survey: identity validation and bot filtering
● In-survey: behavioral and consistency checks
● Post-survey: anomaly detection and data cleaning
This reflects both Section 2 – Primary Data Collection, which governs the use of appropriate methods and quality controls during data collection, and Section 7 – Reporting, which requires transparency in how those methods and controls are applied and communicated.
Extending Fraud Detection Beyond the Survey
An important—and often overlooked—opportunity for improving data integrity emerges after the survey is complete.
Organizations involved in incentive fulfillment and redemption have access to additional behavioral and transactional signals that are not visible during recruitment or fielding. These may include:
● Redemption velocity and patterns
● Reuse of accounts or payment credentials
● Cross-project behavior across incentive programs
● Indicators of coordinated or incentive-driven fraud
These signals can reveal fraudulent activity that was not detectable earlier in the process.
While not explicitly called out in the Code, this approach supports its broader intent—particularly Section 1 – Duty of Care, by taking reasonable steps to identify risks to data quality wherever they emerge, and Section 2 – Primary Data Collection, by extending quality controls beyond the point of survey completion.
Best practice is to evaluate fraud at every available touchpoint, not just at the point of survey completion.
This broader view recognizes a simple reality:
Fraud does not occur at a single moment—it is a process. Detection should be as well.
Real-Time Monitoring and Adaptive Controls
Fraud is dynamic—and so are defenses.
Researchers increasingly monitor data during fieldwork and adapt in real time. Extending that mindset beyond fielding—into fulfillment and redemption—creates a continuous quality loop, rather than a one-time screening event.
This supports both Section 1 – Duty of Care, by ensuring that data quality is actively managed and risks are addressed throughout the research lifecycle, and Section 2 – Primary Data Collection, through the ongoing application and refinement of appropriate data quality controls.
Balancing Detection with Inclusion
A critical ethical consideration is avoiding overcorrection. While catching and eliminating fraud is a critical necessity, eliminating legitimate responses from the data also damages data quality as not all viewpoints may be accurately represented. A balance must be struck.
The Code requires that researchers avoid misleading stakeholders. Overly aggressive fraud removal can:
● Exclude valid respondents
● Introduce bias
● Distort findings
Thoughtful approaches distinguish between signals and proof, ensuring decisions are proportionate and defensible.
Transparency in Fraud Indicators
Transparency is not only about methods—it is also about what the data reveals.
In line with Section 7 – Reporting, researchers should consider sharing:
● The proportion of responses flagged or removed
● Types of fraud detected
● The impact of cleaning and validation on the final dataset
This includes insights gained after fielding, such as fraud detected during incentive redemption or fulfillment.
Bringing It Back to the Code
Fraud prevention is not just a technical function—it is a direct expression of professional standards.
By expanding fraud detection across the full research lifecycle—and being transparent about it—we reinforce core principles of the Insights Association Code of Standards and Ethics:
- Transparency in Methods (Section 2: Primary Data Collection): Clearly communicating how data quality is protected at each stage of the respondent journey, from recruitment through fulfillment.
- Commitment to Sound Methodology (Section 2): Applying rigorous and appropriate techniques—including behavioral and transactional signals—to ensure the validity of collected data.
- Integrity and Data Quality (Section 4): Safeguarding research results so they reflect real, qualified participants rather than fraudulent or duplicated responses.
- Accurate and Responsible Reporting (Section 7): Providing clients with the necessary context about data quality controls to support proper interpretation of findings.
Final Thought
Fraudsters exploit gaps between systems, stages, and stakeholders.
Our response should be equally connected.
The future of data quality is not just better detection—it is continuous, transparent validation across the entire research ecosystem, from recruitment to response to redemption.
That is how we bring the Code to life—and ensure that our insights remain worthy of trust.
This paper has concentrated on fraud that originates outside of the industry. While not frequent, there have been several instances of fraud committed by organizations within the industry. Needless to say, these instances violate the Code as well as the law.
ABOUT THIS SERIES: The Insights Association Code of Standards & Ethics sets the principles that guide ethical and professional market research, insights, and analytics. But how do those standards apply in everyday practice? In this series, members of IA’s Standards Committee bring the Code to life through practical examples, showing how it guides responsible research and decision-making across the industry.