The future of qualitative research: Cookiy AI on the Campaign for Real Qual

On a recent episode of the Campaign for Real Qual, host Craig Griffin, founder and Chief Insight Officer at FUEL ASIA Research and Consulting and a thirty year veteran of the industry, sat down with two people he had met at the IIEX conference in Bangkok: Joven Lee, Cookiy AI's Chief Revenue Officer and President, a market research and go to market leader who spent more than a decade building and scaling large multi market programs across Nielsen, Statista, and Cint, and Davin, Cookiy AI's CEO, who came up through product and engineering at Instagram, TikTok, and Tencent. The show is deliberately impartial, no sponsors, so the invitation was a signal in itself: Griffin thought this team had something genuinely useful to say about where research goes next.

The opening question was blunt. Qual at scale has existed for two or three years and thirty to forty companies now claim to run qualitative research with AI, so what is different here. Davin gave a three part answer: the philosophy is to enhance human researchers, not replace them; the product takes a second mover advantage by building on the 2025 generation of thinking models, real time voice AI, and multimodal understanding rather than older tech; and the design serves both sides of the market, the clients who commission research and the respondents who supply it, with the ambition of becoming the bridge between them. Joven framed the same idea in three words the team keeps returning to: convenience with control. AI can deliver the full scale of convenience, but at every step, discussion guide, recruitment, moderation, the researcher keeps the ability to step in. It is an ecosystem where agencies, freelancers, and research tech coexist, not a replacement story.

Griffin pressed on a point that had stuck with him from their first meeting: Joven had said the team invested heavily in teaching AI how qual actually works. The answer was a nice image. Traditional qual runs behind a one way mirror, where the client and researcher watch the moderator. Cookiy built a second mirror, so the team could watch how moderators moderate and how clients and researchers operate behind the scenes. Serving both sides, Joven argued, exposed the largest gap in the market and became the company's strongest advantage. Davin added the engineering reality underneath: injecting human knowledge into AI while controlling hallucination is hard, a single study runs many sub agents, and the system layers state of the art models, a fine tuned layer, a large proprietary dataset, and an observer and control layer so a human can monitor and intervene.

The most human moment in the conversation was about closeness. Griffin has long worried that qual at scale, insights on demand, frictionless and low effort, risks distancing researchers from what they are listening to, and that a researcher who has not internalized the evidence cannot persuade a client to act on it. Davin agreed from hard experience: at Instagram, Tencent, and TikTok he insisted his own product managers and marketers run interviews themselves rather than delegate everything to agencies. So Cookiy built an instruction mode where a human moderator runs the first couple of sessions, the AI observes and learns, and even once the AI takes over, humans can watch and interrupt in real time. The human stays in the loop with firsthand feeling intact.

On where growth comes from, Joven made two arguments. First, triangulation: the recurring client complaint is that quant, qual, and CRM data do not talk to each other, and the value is in combining them into one decision rather than three disconnected studies. Second, democratization: market research has been an expensive activity, often fifty thousand dollars a project, locked to big brands, when a product manager, an SME, or an individual contributor may want real evidence without that budget. Davin extended it to the age of vibe coding tools like Claude Code and Cursor, where one person can build a company fast and the real moat becomes deciding what to build. An engineer can burn a thousand dollars of tokens overnight building software; spending a few hundred to pinpoint what users actually need first is, as he put it, a very good deal. Griffin connected it to the idea of a billion dollar solopreneur, and to a whole population of builders who would never have hired a traditional research agency.

Asked what stays uniquely human, Davin pointed to non textual signal, the micro expressions, tone shifts, and small gestures that carry meaning. Cookiy's AI tracks faces, tone, and body movement, which he believes is rare in the market, but it is still far from a human expert, so the recommendation is to put researchers back in the observation room for the conversations that matter, with the AI flagging moments a person should interpret. Griffin, based in Southeast Asia for two decades, tied this to high context cultures like Thailand and Japan where much of the meaning sits outside the words, and where human meaning making remains essential.

The closing five year question drew a shared view: the technology will democratize, so durable advantage comes from adding value to both sides of the industry, not from guarding the tech. Joven offered the sharpest advice for practitioners. Research has seen this movie before, with the resistance that met SurveyMonkey, Decipher, Confirmit, and SPSS before everyone adopted them, and the answer now is to be open minded, take some risks, and find the flow that fits your own company, because when AI automates the routine, the ability to identify that flow is the winning ingredient. Otherwise every report starts to look the same and you end up with a copycat identity crisis.

Amplify your originality, not just your efficiency

That point became the natural place to end, and not only because it was well made. Griffin noted it was the exact theme of the keynote he had just delivered at IIEX with David Coffin: in the race to be efficient and drive down cost, use AI to amplify your originality, not just your output. A thirty year veteran hosting an impartial program and a research leader building the next chapter of the field had arrived at the same conclusion, which is about as clean a close as a conversation like this gets.

Craig Griffin hosts the Campaign for Real Qual and leads FUEL ASIA Research and Consulting; find him on LinkedIn at linkedin.com/in/craig-griffin-thailand. Cookiy AI builds AI moderated user research that keeps humans in the loop.