Over the past few weeks, we’ve shared a series of observations from 2025, starting with how AI fraud actually showed up in practice and what PureScore™ 5.0 revealed, followed by a closer look at the myth of the “instant AI survey taker,” and then an examination of where quality can quietly break mid-survey through changes in session continuity.
This final post brings those observations together to reflect on how quality can change, how to notice it, and how to respond with clarity rather than overreaction.
What 2025 highlighted wasn’t a dramatic shift; it was a clearer understanding of how quality can break and what we can do to confidently monitor, flag, and mitigate the risk.
Fraud is a constantly evolving target influenced by incentives, access, and technology. Industry discussions often focus on what respondents say, open-ended questions, logic consistency, and attention checks, but 2025 reminded us to pay closer attention to how respondents navigate a survey. Pre-survey and in-survey checks remain necessary and valuable. What became clearer this year were quality risks originating from changes in the respondent’s environment, not the responses themselves.
Over the course of the year, we saw instances where data quality became suspicious mid-survey, even when initial behavior did not raise risk signals. This raised a broader thought that applies across the research ecosystem: response quality is shaped by the stability of the session. When devices change, when locations shift, or when interaction patterns don’t align with how the session began, the downstream data often reflects that disruption.
As discussed earlier in this series, AI-assisted survey completion still requires human involvement at key moments, creating natural points where a handoff or takeover may occur. Rather than evaluating responses alone, session continuity helps assess whether the survey environment remains consistent from start to finish.
This assumption created an opportunity for us to expand our thinking about quality and how we could more easily assess and address this type of risk. Instead of focusing solely on verifying responses, we can also look at verifying session integrity:
-
- Are the device and environment consistent from survey start to finish?
- Do interaction patterns match what we expect from a single respondent?
- Do we have the signals required to check for session integrity?
The market research industry has always relied on traditional tools such as good survey design, clear logic, and thoughtful in-survey checks. But 2025 reinforced the message that understanding where and why continuity breaks is now also important.
As a company, we have made significant investments in quality initiatives since the launch of PureScore™ in 2019. We continuously release updates, including our most recent PureScore™ 5.0, not because fraud has become more extreme, but because subtle shifts are easy to miss without the infrastructure to monitor them consistently.
We have a corporate commitment to continually identify even small increases in fraud risk and implement fixes to secure better quality data for our customers. Staying ahead of fraudulent activity isn’t always about dramatic interventions; sometimes it’s small signals like device-level clues, location patterns, or interaction patterns that provide the clarity needed to act with confidence.
Let’s carry that clarity forward into 2026.
Mark has nearly 20 years of experience in global business leadership, corporate entrepreneurship, and technology innovation. He has worked passionately within the industry building businesses, tools, and gathering evidence to improve data quality. In his current role as Chief Product Officer of PureSpectrum, he oversees product, engineering, data science, and supply teams across 6 countries and works to deliver new PureSpectrum features that will continue to move the industry forward. Mark graduated from the University of Washington with a double major in Marketing and Corporate Entrepreneurship and added a Certificate in Contract Management from the University of Washington Extension School.
