Article by ICG member Arthur Fletcher
Using Collaborative AI* to produce a research report is, in many ways, like composing a beautiful symphony.
It begins with the interviews themselves: the raw conversations, the stories, the hesitations, the sparks of emotion. These are your musical materials – keys, chords, motifs, riffs and phrases. Each participant offers a new timbre, a different rhythm, a subtle variation on the theme. By the time fieldwork ends, you have an orchestra’s worth of material waiting to be shaped.
For a typical AI analysis, the next step is an analysis plan: an executive summary, a thematic breakdown, key quotes, implications etc. This is immensely valuable but it can feel like a pianola roll – a predetermined script that produces a pleasant, recognisable tune. It follows familiar combinations and harmonic progressions and it gets the job done but it is not yet a symphony. It has form, but not soul.
This is where Collaborative AI changes the creative atmosphere.
The AI takes that pianola roll – the structure, the patterns, the first implications – and lays it out clearly. Then, like a composer settling at the piano, you begin to explore. You start asking questions of your data:
“What happens if I follow this theme?”
“Where does this emotional progression lead?”
“What is the story hiding beneath this discord?”
The AI listens, responds and plays back possibilities. It doesn’t invent notes that were never played, it helps you uncover the melodies the data is already singing. It can isolate a phrase from one interview, harmonise it with a counter-theme from another or spotlight a discord that demands resolution. It allows you to roam through your material with freedom, curiosity and confidence, guided but never constrained.
Slowly, the music becomes richer. Insights that were once scattered become aligned. Subtle tensions resolve into understanding. A fleeting remark becomes the emotional centrepiece of the movement. What began as raw notes and mechanical playback transforms into something shaped by instinct, interpretation and craft.
And then – almost imperceptibly – the final composition emerges.
Not a machine-generated tune, not a formulaic analysis but a symphony – layered, resonant, human in its sensitivity and unmistakably yours.
Collaborative AI does not replace the researcher, it accompanies them. It provides the structure of the composition, the clarity of the score and the responsiveness of an intuitive musical partner. But the symphony – the insight, the narrative arc, the emotional truth – is composed by the human hand.
In this way, Collaborative AI doesn’t diminish the artistry of qualitative research. It amplifies it, offering researchers the chance to move beyond mechanical summaries and into the realm of genuine creative interpretation.
Because when AI and researcher compose together, insight can finally sound as beautiful, as powerful, and as unforgettable as music.
If you would like to know more about Collaborative AI*, please get in touch.
*’Collaborative AI’ is a term I’ve coined to describe an AI that works with the researcher rather than in their place – an intelligence that supports, clarifies and elevates human interpretation while remaining absolutely faithful to the data.
AI also helped me to create this article.
Arthur Fletcher | LinkedIn
In this session, Greg Burke from Tellet will show you exactly how AI-moderated qualitative research works in practice – from designing a discussion guide to analysing hundreds of in-depth conversations at once. You’ll see a live demo of the Tellet platform, hear how independent researchers and agencies are already using it to win new business, and learn how you can start including it in your own proposals.
He’ll also walk through our new referral scheme for ICG members.
Greg Burke is the CEO and co-founder of Tellet. Based in Amsterdam, he has spent the last three years building Tellet into a platform used by some of the world’s leading research agencies and brands, including Kantar, NielsenIQ, Vodafone, and KPMG. Greg is passionate about giving independent researchers access to the same tools the big agencies use.
With 30 years of research experience, Andrew blends qualitative depth with quantitative precision to uncover the insights that drive smarter decisions, stronger customer relationships and business growth.
Jay Zaltzman is based in Palm Springs, California, USA. He is a frequent presenter at research conferences and a past president of the QRCA. He writes an email newsletter called Research Tidbits – short, digestible tidbits about the world of marketing and market research, available here: 
