Insights Teams Warned: AI Can’t Replace Human-Led Qual, Data Quality in 2026 Quirk’s Event

0

Market researchers face a critical balancing act between embracing artificial intelligence and maintaining the human fundamentals of their craft, according to key takeaways from the 2026 Quirk’s Event in New York, where industry leaders warned that the rush toward automation must not come at the expense of empathy, data integrity and actionable insights.

Jill Miller, co-founder of Catapult Insights and a consumer research specialist with decades of experience, recapped three dominant themes from the conference: the enduring power of human-led qualitative research, the evolving data quality crisis, and the persistent bottleneck between insights and action. Brandspur Brand News understands that these themes reflect broader tensions within the industry as researchers navigate an increasingly technology-driven landscape while preserving the core principles that generate trustworthy insights.

Despite the growing attention on AI personas, digital twins and AI-moderated interviews, human-led qualitative work held its ground at the conference, appearing consistently across brands, categories and sessions. Ivey Crespo, VP of qualitative and activation at Vital Findings, and Beth Foster, senior manager of global consumer and shopper insights at The North Face, made the case that while the industry has focused heavily on AI and automation, putting researchers in context alongside the people they are trying to understand remains critical for driving true innovation. A session from ImpactSense and ASICS demonstrated how decades of scientific testing and consumer data still left gaps in understanding comfort that only direct, in-person immersion with runners could close.

Underneath these discussions, immersion is not really about method but about empathy, which can easily become a nice-to-have in a research practice under pressure to move fast and produce more. Yet getting close enough to consumers to understand their context, frustrations and motivations is what gives researchers a holistic view of who they are designing for, and without that, it becomes easy to create products that test well on paper but miss what people actually need.

Data quality emerged as an evolving concern, with Miller’s own session “Clean data, wrong answers: The new data quality crisis in market research” alongside Paul Ponsford of Delta Faucet Company and Alexandrine de Montera of Full Circle Research. The panel argued that most teams have become reasonably good at blocking bots, flagging duplicates and monitoring survey behaviour, but real data quality requires more than technical checks, including consideration of the behavioral aspects of respondents such as inattentiveness and low engagement. Miller noted the resonance of this discussion, with brand-side researchers already encountering these challenges without a shared playbook for addressing them.

Also read: https://brandspurng.com/2026/09/01/south-africans-embrace-ai-but-still-demand-human-customer-support/

The insights-to-action bottleneck also featured prominently, with multiple sessions hitting on the same frustration: well-executed research still gets shelved when researchers have not built the internal relationships or storytelling needed to move stakeholders to action. Reckitt Nutrition’s session reframed the problem as an access issue rather than a reach issue, arguing that most research does not have a reach problem but an access problem, with studies completed and reports filed while insights that should shape decisions never make it to the people who need them.

One brand-side insights leader advised her team to stop working project to project, let AI handle small questions and do their own research to answer bigger questions, reframing where a research team’s attention could be placed to become a more impactful influencer within the organisation. Miller concluded that even as tools multiply, the fundamentals have not gone anywhere: researchers must stay close enough to real people to understand them, know which respondents to trust, and remain disciplined in turning trust into action. Methods will keep evolving, but judgment remains squarely a human job.