POV: You’ve got a line of frozen pizzas that’s doing really well in grocery stores, and you want to add a new option. You’re considering two possibilities: 1) truffle oil with mushrooms; and 2) anchovies with capers.
Truffle mushroom is your obvious pick, but you want to ask your customer base first (as you should). So, you put out a market research survey en masse via social media.
The results come back as a bit of a surprise: 51% pick truffle mushroom, and 49% pick anchovies (by some accounts, the least popular pizza topping ever). Baffled, you decide to offer both.
After spending an exorbitant amount of money to launch two new products, you find that truffle mushroom is insanely popular, but the anchovies one flops. What’s going on? Didn’t the survey show that people liked both almost equally? Clearly, that wasn’t true at all.
You’ve just experienced the damage wrought by survey fraud.
Survey fraud—sometimes known as “panel fraud,” in a nod to consumer panels—occurs when people take advantage of paid surveys in ways that lead to inaccurate data. They might:
It’s all for the purpose of reaping the associated rewards. Many companies offer small financial incentives to garner more participants, but these very incentives encourage deceptive behavior.
It’s pretty bad news for you or any other company relying on accurate, population-representative data for market research. You’re in a Catch-22, where stopping the compensation will lead to fewer people taking your surveys, but continuing to offer it attracts the bad actors and floods your research with nonsensical data.
(Note that we’re not talking about malicious surveys created by scammers to steal people’s PII, or personally identifiable information. That’s a big problem too, but not one that impacts product testing research.)
If your survey results are full of garbage data, it could really tank your next product launch. You may find yourself betting big bucks on the wrong strategy, product, or campaign.
Dishonest and duplicate answers poison the consumer insights you’re intending to use as a basis for major product launch decisions. You might end up worse off than if you’d simply gone with your own gut feeling.
Cleaning up the data isn’t cheap either. It can be tough to tell which responses are accurate and which should be tossed out—particularly when respondents number in the hundreds or thousands.
Essentially, your budget is being wasted by a new, illegitimate “profession” that’s been spun up around one of the most important tools of our era in market research. This is why survey fraud prevention is so critical.
Since the early 2000s, there’s been a segment of the population that tries to make some side money by completing mundane tasks online, usually via platforms like Swagbucks or Mechanical Turk. But then in 2020, while the COVID-19 lockdowns kept people at home, the ranks of paid survey-takers swelled.
It’s important to note that most of these people are NOT trying to commit fraud. Many of them are retired, on disability, or in a caregiving role, and they’re just looking to make an extra $30-40 a month. They answer questions honestly and take only the surveys they qualify for.
However, the rosy picture that shows up in blog posts with titles like “Top 15 Side Hustles in 2025” misrepresents the nature of survey-taking. Participants usually get far, far below minimum wage, sometimes just pennies on the hour, and they often don’t get paid anything at all to attempt surveys that ultimately disqualify them after a whole page of questions. So, you can see how someone might be motivated to game the system.
This is particularly true if this person is tech-savvy and knows how to whip up a network of bots that will take multiple surveys at once while hiding their IP address with a VPN or a proxy server.
The result of all this? Brands seeking legitimate consumer insights are up against increasingly sophisticated fraud.
It’s not easy. Survey fraud detection requires creativity, tech-savviness, and multiple layers of defense.
You can start by weeding out bots and fake email addresses, then catch dishonest (or non-human) participants with specialized types of questions. Wrap it all up by analyzing participant behavior and disqualifying those who don’t act like a normal, well-intentioned respondent.
Here’s a list of some things you can do for survey fraud prevention:
As always, make sure that you’re collecting data responsibly and respecting users’ privacy, even when including certain questions for the sole purpose of bot detection.
Remember when most CAPTCHAs were a bunch of squiggly letters, and you had to type in what you thought they were? Bots started reading them with optical character recognition (OCR). As a result, the letters got squigglier and squigglier—to the point where even humans could barely read them at all. And OCR just got better and better.
Eventually, those CAPTCHAs got replaced with a new kind that analyzes a person’s mouse movements, browsing history, and browser cookies while they check a box. Unfortunately, the bots are also finding ways to simulate human-like mouse movements, and they’re grabbing cookies via web scraping.
This is just one example of how, for every fix, the fraudsters soon have a new workaround. They’re ingenious, and they’re persistent. This is their livelihood.
Survey companies that don’t vet their audiences beforehand are always going to be besieged by this. It’s really hard to trust the data when you know that people are using AI to generate plausible-sounding responses and finding increasingly clever ways to make it look like all their bots are actually hundreds of individual human users.
At Highlight, we don’t want to be constantly trying to root out bots and dishonest survey-takers. That’s why we focus on cultivating an audience of committed testers who we know will give accurate, thoughtful feedback every time.
We don’t need to include challenge questions or honeypots. This is because we don’t admit people who are trying to game the system.
Rather than sinking a bunch of your precious mental effort (and money) into the problem of “how to detect survey fraud,” you can simply focus on writing good market research questions that get to the heart of what you need to know.
The future lies with vetted product testing communities, and that’s what Highlight has built. When you work with people who have already demonstrated their ability to offer valuable, nuanced insights, you can bypass the world of bots altogether.
With the technologies out there today, you can expect survey fraud to only get worse. But a product testing community like Highlight’s will continue to grow and thrive, helping brands make ever-smarter decisions about their next launch.