Data Gurus podcast: Solving data quality challenges panel

In the latest episode of the Data Gurus Podcast, host Sima Vasa gathered a panel of experts to tackle one of the most pressing issues in research today: solving data quality challenges. Panelists included Bob Fawson of Data Quality Co-op, Dyna Boen of Escalent, John Kay of Intuit, and Rep Data’s Steven Snell, Phd,  our EVP and Head of Research. Each shared perspectives on why quality problems persist, and what it will take to fix them.

Steve reminded listeners that data quality isn’t just about catching fraud. “Data quality is a function of at least two things—fraud and inattention,” he said. “We’ve been worried about inattentive respondents for decades, but the tools we use to identify fraud still look like we’re checking for speeders and straight-liners. They don’t catch the sophisticated, ‘good-looking fraud’ that’s quietly reshaping our data.”

Understanding “good-looking fraud”

Unlike traditional bad data, “good-looking fraud” passes basic quality checks. These responses often sound human and seem thoughtful, but subtle inconsistencies reveal their artificial or duplicate origins. “You look at it and think, that’s good—but it’s too good,” Steve noted during the podcast panel. These polished yet false responses distort KPIs and undermine trust in results if not detected early.

A layered approach to quality

During the Data Gurus podcast, Steve emphasized that solving data quality challenges requires a lifecycle mindset:

• Before fielding – Build protection in early with robust fraud prevention and passive data checks
• During the survey – Keep respondents engaged and attentive without exhausting them through overly complex validation tasks.
• After fielding – Clean and validate datasets to ensure only trustworthy responses remain.

“At Rep Data, we think about all of these signals together,” Steve explained. “We look at who these people are, where they’re coming from, and what technology they’re using. One of the things we monitor is how many times we’ve seen a respondent in the Research Defender ecosystem in the past 24 hours.”

Protecting both data and respondents

Steve cautioned that while tighter controls are essential, researchers must balance rigor with empathy. “We have to be careful because respondent experience matters. If we give survey respondents a bad experience, even the good people are going to give us bad data.”

Takeaway for researchers

“Data quality begins before the survey and ends after it,” Steve concluded. “It’s not just about fraud or inattention—it’s about their intersection and where they don’t overlap. That’s going to require a researcher’s attention across the full life cycle.”

Listen to the full Data Gurus Podcast episode, “Solving Data Quality Challenges,” featuring Steve Snell of Rep Data, Dyna Boen of Escalent, Bob Fawson of the Data Quality Co-op, and John Kay of Intuit: https://www.infinity-2.com/podcasts/solving-data-quality-challenges-with-dyna-boen-of-escalent-jon-kay-of-intuit-bob-fawson-of-data-quality-co-op-and-steven-snell-of-rep-data/