Online research fraud has boomed since the onset of the pandemic. Kantar’s Profiles division estimates that industry wide, 15 million surveys have been “completed” by international hackers in 2023 and more than 9 million completed already in 2024 – a 780x increase since 2018.
Not addressing this fraud can have a serious impact on data quality, consistency and all-around reliability. The data below, for example, shows the difference in data between international hackers and genuine US respondents. This level of overclaiming could result in entirely wrong business decisions being made.
Threats to data integrity can come from a range of bad actors - from the very damaging hackers and bots to less severe disengaged/fatigued respondents. Here are three valuable focus areas researchers should seek out before choosing an online research partner.
1. Know your respondent sources
Understanding where your respondents come from is the first step in enhancing data quality. Not all panels and respondent sources are created equal. High-quality panels invest heavily in their panellist experience and technology, treating respondents as valuable contributors rather than mere commodities. This investment in quality leads to more accurate and reliable data, which is essential for making informed business decisions.
Panels vs. Exchanges
High-quality panels, like those maintained by Kantar Profiles, invest significantly in their acquisition sources, panellists' experience and technology. This approach ensures that respondents are genuine, engaged, and representative of the target population. These panels use robust recruitment processes, rigorous screening, anti-fraud technologies and continuous engagement strategies to maintain a high standard of data integrity. In contrast, exchanges might treat respondents as interchangeable commodities in volume, leading to lower engagement and higher rates of fraud and disengagement, which can compromise data quality but can deliver at speed and scale.
2. Optimise the experience for respondents
A positive respondent experience is crucial for obtaining high-quality data. Happy and engaged respondents spend more time in studies and provide more accurate and thoughtful answers, which leads to better data quality and more reliable insights.
Optimise through survey design
Design surveys respondents want to take, ensuring they are user-friendly and optimised for mobile devices. Given more than 80% of panellists join panels via mobile devices, it is essential to create mobile-friendly designs that are easy to navigate and complete. This includes using clear and concise language, avoiding overly complex questions, and ensuring that the survey layout is visually appealing and accessible on smaller screens.
Consider the questions themselves as well and whether or not you’re approaching your designs with empathy. An empathetic approach is one where potential judgement is removed, allowing respondents to answer open and honesty, thus improving the quality of insights uncovered.
Optimise through sourcing
The panel landscape is rapidly evolving, and with the world becoming increasingly digital, how panel managers engage with panellists must evolve, too. When high-quality respondents matter most to your research, like with longitudinal projects, find a data partner that addresses these changes. At Kantar Profiles, for example, we use machine learning algorithms that analyse panellist behaviour and get the right person into the right survey at the right time for optimal fieldwork efficiency and respondent engagement.
3. Invest in quality
Investing in data quality is non-negotiable. Cutting corners here can lead to significant issues down the line, including skewed data and unreliable insights that can turn into poor decision making and missed opportunities. High-quality data is the foundation for accurate analysis and informed decision-making.
Fraud prevention
Fraud is sophisticated and ever-evolving. There is a cost associated with identifying and excluding fraudulent respondents before they enter your survey. Kantar Profiles uses advanced fraud detection systems that leverage machine learning and other technologies to identify suspicious behaviour patterns and anomalies in real-time. These systems can detect fraudulent activity at multiple stages of the data collection process, from initial recruitment to survey completion. If you're buying cheap, expect less reliable data and more time spent weeding through responses. Investing in panel that uses robust fraud prevention measures is essential to maintain the integrity of your data and ensure that your insights are based on genuine and accurate responses.
Learn more about preventing survey fraud, here.
Quality as a foundation
Quality is foundational, not discretionary. Low-quality sources can cause significant data shifts and skews that cannot be corrected and identifying the samples that caused the skew becomes a tedious subjective exercise. Inconsistent or unreliable data can lead to incorrect conclusions and misguided business strategies.
Investing in high-quality data sources from the outset prevents these issues and ensures reliable, actionable insights. This involves selecting research partners with proven track records of data quality, implementing comprehensive quality control processes, and continuously monitoring and improving data collection methods. By prioritising quality at every stage of the research process, you can ensure that your data is trustworthy, and your decisions are well-informed.
Kantar's commitment to data quality
At Kantar Profiles, we have devoted significant resources to building confidence in our data collection processes. We have extensively researched panellist quality and invested in advanced technology to detect poor behaviour immediately. By focusing on the panellist experience and optimising survey delivery, we ensure that we get the right people into the right surveys at the right time, asking the right questions to deliver highly engaged answers. Here are our key areas of focus:
Global panel quality & global reach
Kantar Profiles' global panel quality is a cornerstone of our commitment to data integrity. Our six Premium Panels are comprised of genuinely engaged panellists across the globe, covering 54 markets and 26 languages. Our investment into our panel brands drives meaningful connections with diverse groups of people who matter most to our clients’.
- Detailed panellist profiling: We deeply profile our panels to understand the unique attributes of each panellist. This enables us to target surveys to the right audiences, increasing the relevance and accuracy of responses.
- Continuous panel engagement: To maintain high engagement levels and data reliability, we create a community people want to be a part of. Not only do we pay our panellist appropriately for their time and data, we share insights, connect with them regularly, and ensure they feel valued.
Our Profiles Respondent Hub, an API-enabled single access point for our Premium Panels and trusted partners, delivers the quality of data you expect from Kantar with even more reach and scale. With access to over 170 million people across 100 markets, our clients experience fast, reliable, high-quality global data at scale – all in one place.
Industry-Leading Fraud Prevention
In an era where online research fraud is rampant, Kantar Profiles' industry-leading fraud prevention measures are crucial. Our proprietary, industry-leading AI and machine learning anti-fraud technology, Qubed AI, identifies fraud within milliseconds. It also helps train lower-quality panellists to become meaningful contributors.
- Advanced Machine Learning Algorithms: We leverage advanced machine learning algorithms and years of labelled data points to train, monitor and analyse panellist behaviour in real-time. These algorithms are trained to detect anomalies and patterns indicative of fraudulent activity, ensuring only high-quality data is collected.
- Biometric verification: Our multi-layered verification process includes facial evaluations. Using AI video analysis, it detects duplicate respondents and validates demographics at registration as need. This ensures that only real, unique people join our panels.
- Open-ended response evaluation: Using genernative AI (LLMs) technology, we score open-ended answers on relevancy, originality, completeness and language. We use this catch fraudulent behaviour and support the quality of our panellists.
- Continuous fraud monitoring: Fraud prevention at Kantar Profiles is an ongoing research and development process. We continuously update our fraud detection systems to counter new threats and techniques used by fraudsters, keeping our data collection processes secure and reliable.
Expertise in survey design
Our expertise in survey design is another critical aspect of our commitment to data quality. We help clients ask the right questions at the right time in the right way to avoid biases and dishonest answers, ensuring the data collected is both accurate and actionable.
- Empathetic and mobile-optimised design: Understanding that a positive respondent experience leads to better data, our survey design best practices are empathetic and centre around mobile respondents. This means supporting surveys that are easy to navigate on all devices, reduce respondent fatigue, elicit honest answers, and improving engagement.
- Flexible survey support and engagement techniques: We work with clients to design research for their specific audience and research goals. This collaboration allows us to tailor our data collection to meet specific objectives, delivering data that drives impactful business decisions. Whether it’s using gamification to increase engagement or employing open-ended questions for deeper insights, our methodologies are also designed to elicit high-quality responses.
- Data-driven improvements: Our approach to data quality is iterative. By conducting ongoing research-on-research, we continually refine our best practices to enhance data quality. This data-driven approach ensures that our methodologies evolve with the changing landscape of online research
Kantar Profiles is the industry leader in data quality. By investing in innovations to drive real change in the industry with best-in-class self-learning adaptive technology, Kantar is the best home for any research program. For a more in-depth walkthrough of how Kantar continues to tackle online research fraud, and how you can protect your data, watch the on-demand Webinar, “Everything you need to know for collecting the highest quality data.”
