Ask your interview data questions and get answers with quotes

A stakeholder reads your report and asks why price came up so often among lapsed buyers. Answering means reopening dozens of transcripts, searching for every mention and copying out the quotes that back it up. Questions like that often go unanswered, because the digging takes longer than the meeting they came up in. Talk to Your Data lets you ask the study instead.

Talk to Your Data, how it works and what it changes, no sound needed.

How it works

• Ask in plain language. Open Talk to Your Data from a study and type your question, for example "What are the main reasons people choose a competitor?"

• Answers draw on every interview. The AI research assistant works across the transcripts and answers with the supporting quotes and the participants who said them.

• Same idea, different words. It connects participants who describe the same thing in different language, which a keyword search would miss.

• Keep digging. Follow up to narrow a segment, test a hypothesis or ask for more evidence on a single point.

Build on research you already ran

Earlier studies stay useful. Before you write the next topic guide, ask what past research said about a segment or a claim, and start from what you already know.

Research teams asked for it in five situations

These come from teams we work with in consumer goods, food and drink, beverages and spirits, entertainment, travel, software and research agencies. We kept them anonymous.

• Follow-up questions after the report. When the readout raises something new, teams wanted to ask the study directly and get an answer the same day.

• The evidence behind a finding. Before a line goes into the deck, researchers wanted the raw responses and the moments behind it.

• Across studies, markets and time. Teams asked to question a brand's research as a whole: several countries, several waves, users and non-users side by side.

• Stakeholders with their own questions. Agencies and brand teams wanted clients and colleagues to put their own questions to a study.

• An early read during fieldwork. Teams wanted to check the direction after the first interviews and decide whether another wave was worth running.

What this means for your research

Follow-up questions get answered while they still matter, with the evidence attached. Time that went into hunting through transcripts goes into deciding what the findings mean.

Some limits to know. The assistant answers only from the interviews in the studies you give it. Open the quotes behind a conclusion before it goes into a deck, and ask it to say how many participants a count refers to. Usage counts toward a monthly allowance where usage billing is on; see usage and limits.

This entry covers the first release, from August 2025. The chat has since been rebuilt: answers stream in, source cards link each finding to the original interview, and the assistant can suggest more studies to add mid-conversation. Read the Talk to Your Data guide.