You’ve spent weeks crafting the perfect survey. You’ve refined every question, polished the logic, and launched it to your target audience, only to face the most demoralizing task in research: cleaning the data. You find yourself deleting responses from bots, filtering out duplicates, and trying to make sense of nonsensical answers from disengaged participants. It’s a frustrating, time-consuming process that can make you question the value of the entire effort. It doesn’t have to be this way. Protecting your data quality goes beyond playing whack-a-mole with bad actors. This guide will give you the strategies you need to fix survey fraud and get the clean, reliable data you need.
In an era where survey results fuel $1 trillion in business decisions, ensuring the integrity of survey data is incredibly important. But the rise of generative AI and increasingly sophisticated bots combined with incentives for bad actors has led to a significant decline in survey quality.
Kantar research indicates that up to 38% of collected data is currently discarded due to issues such as low quality and duplicate respondents.
VerifEye technology uses lightweight facial verification to fight bots and tackle user fraud…
This costly problem led Realeyes to create its Verify technology, which uses lightweight facial verification to fight bots and tackle user fraud dramatically better than digital fingerprinting, CAPTCHA and other solutions. It is a plug-and-play face verification system that’s easily integrated into the market research workflow as a “survey gate” through which all respondents must first pass. VerifEye ensures that your market research remains untainted by fake engagement, providing you with accurate and quality sample data.
We were proud to conduct a proof of concept with Kantar Profiles in Q4 2023, which identified 96.2% of duplicate respondents, bots, and hacking with only a 0.67% false positive rate on duplicates.
We’re even prouder to announce an extensive partnership where Kantar Profiles will embed Verify into its industry-leading anti-fraud software, QubedAI.
With this integration, Kantar’s clients will benefit from:
- Selective respondent validation pre-survey for Kantar Panels and third-party sources
- Third-party panel duplicate respondent protection for client research
- Reduced usage of IP de-duplication rules, which negatively affect mobile IPs
- QubedAI training tool for validating decisions
- Account protection for Kantar-owned LifePoints and Qmee panelists
The integration of VerifEye into QubedAI represents a significant step forward in the fight against survey fraud. With Kantar’s global reach and Realeyes’ cutting-edge technology, the stage is set to revolutionize the way organizations make business and marketing decisions, ensuring that data remains a trusted and valuable asset in an increasingly data-driven world.
What Is Survey Fraud?
At its core, survey fraud is what happens when you get dishonest or fake answers to your survey questions. This isn’t always a case of someone trying to maliciously sabotage your research. Sometimes, it’s a participant who’s just rushing through to get to the end. Other times, it’s a more deliberate attempt to cheat the system for a reward. Regardless of the intent, the outcome is the same: the data becomes unreliable, and any decisions you base on it are built on a shaky foundation. When you’re trying to understand your customers or test a new product, this kind of bad data can lead you in the completely wrong direction, costing you time and money.
Types of Malicious Survey Fraud
While some bad data comes from simple inattention, a significant portion is the result of deliberate, malicious fraud. These bad actors use various tactics to exploit survey systems, usually for financial gain. Understanding these different types of fraud is the first step in building a defense against them. It helps you recognize the patterns and vulnerabilities in your own data collection process. From automated programs to people creating multiple accounts, each method presents a unique challenge to data integrity. Let’s break down some of the most common culprits you’re likely to encounter in the field.
Reward Hunters
You’ve probably seen them before. Reward hunters are individuals motivated solely by the incentive offered for completing a survey. They have no genuine interest in providing thoughtful feedback. Instead, they speed through questions, selecting random answers or providing the bare minimum in open-ended fields just to reach the finish line and claim their reward. Their goal is volume, so they complete as many surveys as possible, as quickly as possible, leaving a trail of low-quality, useless data in their wake.
Automated Bots
Automated bots are one of the most pervasive threats to online surveys. These are not real people but software programs designed to fill out surveys at a massive scale. They can complete hundreds or thousands of forms in the time it takes a single person to finish one, flooding your dataset with completely fabricated information. Because they aren’t human, they can’t provide genuine opinions or experiences, making their responses worthless. Detecting these bots is critical, as they can dramatically skew results and make your entire research effort invalid.
Multiple Submissions
This type of fraud occurs when a single person or a coordinated group submits survey responses multiple times while appearing to be different individuals. They might use various email addresses, IP addresses, or devices to bypass basic security checks. This tactic makes your sample size seem larger and more diverse than it actually is, but in reality, you’re just hearing the same voice over and over again. It fundamentally undermines the principle of gathering unique perspectives from a representative audience.
Fake Answers
Fake answers are exactly what they sound like: responses that are intentionally false or nonsensical. This can range from a participant typing gibberish into a text box to someone fabricating details about their life and habits to qualify for a survey they otherwise wouldn’t be eligible for. While sometimes easy to spot, sophisticated fake answers can be harder to detect, subtly contaminating your data and leading to flawed conclusions about your target audience’s true behaviors and preferences.
Non-Target Audience Respondents
Your survey is designed for a specific group of people—moms with young children, software developers, or frequent travelers, for example. Non-target audience respondents are people who don’t fit your criteria but participate anyway, usually to get the reward. They might lie on screener questions to qualify for the study. This is a huge problem because their feedback isn’t relevant to your research goals. You end up making decisions based on input from the wrong people, which can lead to misguided product features or marketing campaigns.
The Human Cause of Bad Data
Beyond the world of bots and malicious fraudsters, there’s a more subtle, and arguably more widespread, threat to data quality: real, well-intentioned humans. It’s easy to focus on catching the cheaters, but what about the genuine participants who are simply bored, tired, or disengaged? The design of the survey itself can often be the culprit. When faced with long, repetitive, or confusing questionnaires, even the most honest respondent can experience fatigue. This leads to “satisficing”—a fancy term for doing the bare minimum. They might start speeding through questions or choosing neutral answers just to get it over with, which is just as damaging as intentionally fake data.
How Participant Fatigue Undermines Data Quality
Participant fatigue is a silent killer of data quality. It sets in when a survey is too demanding, causing the respondent’s attention and motivation to drop off a cliff. According to research from Rival Tech, this kind of boredom and fatigue can be an even bigger threat than outright fraud. Even after you’ve filtered out all the bots and duplicates, you can be left with a dataset filled with low-effort responses from real people who simply checked out mentally halfway through. This results in straight-lining (choosing the same answer for every question), random responses, and incomplete open-ended answers. Ultimately, this poor experience for the participant translates directly into poor quality insights for your business.
Common Methods for Detecting Fraud
When bad data makes its way into your research, the first step is identifying it. Several traditional methods can help flag suspicious activity, though they often require manual review and can be outsmarted by more advanced bots. These techniques serve as a foundational layer of defense, helping you catch the most obvious offenders. Think of them as the basic security checks before you bring in more advanced technology. While they aren’t foolproof, they can significantly reduce the amount of fraudulent data that pollutes your results by catching inconsistencies, illogical behavior, and other red flags that signal a respondent isn’t who—or what—they claim to be.
Attention and Consistency Checks
One of the simplest ways to spot a fraudulent or inattentive respondent is by checking for consistency. This involves asking for the same piece of information in different ways at various points in the survey. For example, you might ask for a respondent’s birth year at the beginning and their age near the end. If the two answers don’t align, it’s a strong indicator that the person is not paying attention or is a bot providing random answers. These internal consistency checks act as small tripwires, designed to catch those who are rushing through without reading the questions carefully, helping you filter out low-quality responses.
Honeypot Questions
Honeypot questions are clever traps designed specifically to catch automated bots. These questions are typically hidden from human view using CSS or other web technologies. A human respondent will never see the question, so they won’t answer it. A bot, on the other hand, reads the page’s code and will likely fill in an answer for the hidden field. When you see a response to a honeypot question, you have a clear signal that you’re dealing with a non-human participant. This method is highly effective for weeding out simple bots that scrape and fill forms without rendering the page as a human would.
Checking Completion Time
The time a respondent takes to complete a survey is another critical data point. Every survey has an expected range for completion time based on its length and complexity. If a participant finishes a 15-minute survey in under two minutes, it’s a massive red flag. This behavior, known as “speeding,” suggests the respondent clicked through without reading or comprehending the questions. Analyzing these “time-to-complete” metrics allows you to identify outliers who are clearly not providing thoughtful answers, whether they are bots programmed for speed or humans just looking for a quick reward.
IP Address Tracking
Monitoring the IP addresses of your survey respondents can help you detect certain types of fraud. If you receive a large number of submissions from the same IP address, it could indicate a single person is trying to take the survey multiple times or that a bot farm is at work. However, this method requires careful handling. A shared IP address could also come from a university, a large corporation, or a public Wi-Fi network. Furthermore, relying too heavily on IP tracking can unfairly penalize mobile users, whose IP addresses can change frequently. It’s a useful tool, but one that works best in conjunction with other fraud detection techniques.
Strategies to Prevent Survey Fraud
While detecting fraud after the fact is necessary, preventing it from happening in the first place is far more effective. By implementing proactive strategies, you can create a more secure survey environment that discourages bad actors and ensures the data you collect is from genuine, engaged participants. These preventative measures act as a gatekeeper, making it much harder for bots and fraudulent users to access your survey and compromise your data. From controlling access to structuring your questions thoughtfully, these steps help build a strong defense against the most common forms of survey fraud, saving you time and resources down the line.
Use Access Control and Unique Links
One of the most effective ways to stop multiple submissions from a single individual is to control who can access your survey. Instead of using an open link that anyone can click, provide participants with unique, single-use links. This ensures that each person can only take the survey once. For panel-based research, requiring a login to a secure portal before accessing the survey adds another layer of protection. This approach not only prevents duplicate entries from “reward hunters” but also helps maintain the integrity of your sample by ensuring responses come only from your invited and vetted participants.
Randomize Questions and Answers
Automated bots are often programmed to follow a specific path through a survey. You can disrupt their patterns by randomizing the order of questions and the answer choices for multiple-choice questions. When the layout changes for each participant, it forces a level of engagement that bots can’t replicate. This simple technique makes it significantly harder for a script to navigate the survey correctly. For human respondents, it also helps reduce order bias, where the sequence of questions can unintentionally influence their answers, leading to more accurate and reliable data overall.
Design Smart Reward Systems
The way you structure your incentives can either encourage or discourage fraudulent behavior. If you offer a flat reward simply for completion, you invite people to rush through as quickly as possible. Instead, design a reward system that values thoughtful and high-quality responses. This could involve offering bonuses for well-written open-ended answers or using a tiered system where more engaged participants earn greater rewards. By aligning your incentives with the quality of data you want to receive, you can attract more conscientious participants and deter those who are only interested in a fast payout.
Ask Pre-Screening Questions
Before a participant even begins the main survey, use a series of pre-screening questions to ensure they fit your target demographic. These questions help you filter out respondents who don’t meet your specific criteria, such as age, location, or purchasing habits. This not only improves the relevance of your data but also acts as another barrier to bots and fraudsters who may not be programmed to answer nuanced screening questions correctly. A well-designed screener is your first line of defense in making sure you’re talking to the right people for your research.
Provide Clear Rules for Participants
Never underestimate the power of clear communication. At the beginning of your survey, take a moment to explain the importance of honest and thoughtful answers. Let participants know that their genuine feedback is valuable and will contribute to meaningful research. You can also briefly state that the data will be checked for quality and that fraudulent or rushed responses will be discarded without reward. Setting these expectations upfront can encourage legitimate participants to be more diligent and can deter potential fraudsters who realize their low-effort attempts will be caught.
How to Spot a Survey Scam: A Guide for Participants
While companies work to fight survey fraud, participants also need to be vigilant. Scammers often create fake surveys to steal personal information or money. Knowing the warning signs can help you protect yourself and avoid falling victim to these schemes. Legitimate market research companies value your time and privacy, and their practices will reflect that. A survey that feels unprofessional, asks for sensitive data, or makes promises that seem too good to be true is likely a scam. By learning to recognize these red flags, you can confidently participate in real research while steering clear of fraudulent traps.
Warning Signs of a Scam
Scammers often give themselves away if you know what to look for. The most obvious sign of a scam is a request for highly sensitive information that has no place in a legitimate market research survey. They might also use the lure of an impossibly large reward to get you to click a malicious link or give up your data. A real research opportunity will be transparent and professional, whereas a scam will often feel urgent, unprofessional, or overly generous. Trust your instincts—if something feels off, it probably is.
Asking for Payment or Sensitive Information
Here’s a golden rule: legitimate survey companies pay you, not the other way around. If a survey asks you to pay a fee to participate or to access a list of high-paying surveys, it’s a scam. Similarly, you should never be asked for your Social Security number, full birth date, or bank account details. While some survey panels might pay via PayPal and need your email address, they will never ask for your password or direct banking information within a survey. Any request for this type of sensitive information is a major red flag.
Promising Unrealistic Pay
Market research surveys typically offer modest compensation for your time, like a few dollars or points toward a gift card. If you see an offer promising $100 for a five-minute survey, you should be extremely skeptical. Scammers use these unrealistic payouts as bait to lure people into their traps. They know that the promise of easy money is a powerful motivator. Remember that if an offer sounds too good to be true, it almost certainly is. Stick with reputable survey platforms that offer reasonable and consistent rewards.
Requiring Software Downloads
A legitimate survey will almost always be conducted entirely within your web browser. If a survey prompts you to download a special piece of software or an application to participate, close the window immediately. These downloads are often a Trojan horse for malware, spyware, or viruses that can infect your device, steal your personal information, or damage your files. There is no valid reason for a standard survey to require a software installation, so treat any such request as a serious security threat.
Setting Outrageous Payment Minimums
Another common tactic used by scam survey sites is to set an extremely high payment threshold. They might let you accumulate earnings, but when you try to cash out, you discover you need to reach an impossibly high minimum, like $100 or more. You end up completing dozens of surveys, giving away your data for free, only to find that you can never actually reach the payout amount. Reputable companies have low and easily attainable payment minimums, often as little as $5 or $10.
Use the Four P’s to Spot Fraud
When you encounter an unsolicited email or ad for a survey, you can use a simple framework to assess its legitimacy. The Federal Trade Commission recommends a method that is easy to remember: the Four P’s. This approach helps you quickly identify the common tactics scammers use to trick people. By looking for these four elements—Pretend, Problem, Pressure, and Pay—you can become much better at spotting and avoiding fraudulent requests, whether they come in the form of a survey, email, or phone call.
Pretend
Scammers often impersonate organizations you know and trust. They might use the logo of a well-known brand or claim to be from a government agency to gain your confidence. They are counting on you to trust the familiar name and not look too closely at the details, like the sender’s email address or the website’s URL. Always be skeptical of unsolicited communications and verify the sender’s identity through an official channel if you have any doubts.
Problem
The next step in a scam is to create a sense of urgency by inventing a problem. They might claim your account has been compromised, you owe money, or you’re about to miss out on a fantastic opportunity. This fabricated problem is designed to make you anxious and bypass your critical thinking. They want you to react emotionally rather than logically. A legitimate company will not use threats or high-stakes drama to get you to participate in a survey.
Pressure
Once they’ve presented the problem, scammers will pressure you to act immediately. They’ll use phrases like “act now,” “offer expires in one hour,” or “your account will be suspended if you don’t respond.” This pressure is designed to prevent you from taking the time to think, do some research, or talk to someone about the request. They know that if you pause and consider the situation, you’re more likely to realize it’s a scam. Legitimate opportunities don’t rely on high-pressure tactics.
Pay
Finally, scammers will tell you how to pay them, and it’s almost always in a specific, unusual way. They might demand payment via gift cards, wire transfers, or cryptocurrency. These payment methods are difficult to trace and nearly impossible to reverse, which is why criminals prefer them. Remember, if you’re the one being asked to pay for anything related to a survey, it’s a scam. You are the one providing a service with your time and opinions, so you should be the one getting paid.
Ready to Fix Survey Fraud?
Please leave your details if you’re interested in removing user fraud.
Frequently Asked Questions
My current fraud detection methods catch some bad data. Is that not good enough? It’s great that you have systems in place, and they absolutely help catch the most obvious offenders. The challenge is that these methods often miss the more sophisticated fraud that’s becoming common. Think of it like this: you might catch someone speeding through questions, but what about an AI-powered bot designed to mimic human behavior? Research shows that even with traditional checks, a huge percentage of data is still being thrown out. The goal isn’t just to catch some fraud, but to build a dataset you can trust completely, which requires a more proactive and powerful line of defense.
You mentioned participant fatigue. How can technology solve a human problem like boredom? That’s a fair question. Technology can’t force a tired participant to be more engaged, but it can solve a critical first problem: ensuring the person is a real, unique human to begin with. By verifying every respondent at the start, you separate the malicious fraud (bots, duplicates) from the low-quality human data (fatigue, inattention). This allows you to focus your survey design efforts on keeping real people engaged, knowing you’ve already eliminated the noise from non-human sources.
Isn’t randomizing questions and using unique links enough to stop bots? These are smart, foundational strategies that definitely make a difference, especially against simpler bots. However, today’s fraudulent programs are far more advanced. They can often adapt to randomized patterns and find ways around single-use links. Relying only on these techniques puts you in a constant cat-and-mouse game where you’re always reacting to the last threat. A better approach is to stop them at the door before they even get a chance to see your questions.
We’re worried about participant privacy. How does facial verification work without collecting sensitive personal data? This is a super important concern, and one we take seriously. Modern facial verification technology isn’t about identifying who someone is. Systems like VerifEye perform a quick, lightweight check to confirm two things: that there is a real, live person in front of the camera, and that this specific person hasn’t taken the survey before. It creates a mathematical representation of the face to check for uniqueness, but it doesn’t store the image or connect it to any personal identity. It’s about verifying human presence, not tracking individuals.
Can I really trust the data even after filtering out fraud? Confidence is the ultimate goal. The problem with only detecting fraud after the fact is that you’re left wondering what you might have missed. By using a verification gate at the very beginning of the process, you ensure that your starting pool of respondents is made up of real, unique individuals. This fundamentally changes the quality of the data you collect. When you know your responses are coming from your actual target audience, you can have much greater confidence in the insights and the business decisions you base on them.
Key Takeaways
- Fight Bad Data on Two Fronts: Don’t just focus on malicious bots and fraudsters. Disengaged human participants who rush through questions can harm your data quality just as much. A complete strategy involves both blocking bad actors and designing surveys that keep real people focused.
- Prioritize Prevention Over Cleanup: Filtering bad data after you’ve collected it is a time-consuming and frustrating task. Get ahead of the problem by using unique survey links, randomizing questions, and creating smarter incentives that reward thoughtful responses, not just speed.
- Confirm You’re Hearing from Real People: As bots become more sophisticated, traditional checks like IP tracking and honeypot questions are no longer enough. Using modern tools like facial verification is the most reliable way to ensure a real, unique human is behind every response, protecting the integrity of your research.