---
title: "Qualitative Research at Scale: Methods, Tools and… | Minds"
canonical_url: "https://getminds.ai/use-cases/qualitative-research-at-scale"
last_updated: "2026-10-04T20:54:34.914Z"
meta:
  description: "How teams run qualitative research at scale in 2026: AI-moderated interviews, synthetic Audiences and AI-assisted coding compared, with costs, limits and a..."
  "og:description": "How teams run qualitative research at scale in 2026: AI-moderated interviews, synthetic Audiences and AI-assisted coding compared, with costs, limits and a..."
  "og:title": "Qualitative Research at Scale: Methods, Tools and… | Minds"
  "twitter:description": "How teams run qualitative research at scale in 2026: AI-moderated interviews, synthetic Audiences and AI-assisted coding compared, with costs, limits and a..."
  "twitter:title": "Qualitative Research at Scale: Methods, Tools and… | Minds"
---

Minds

June 12, 2026·Updated October 3, 2026·Use-case·Jerry Miller, Product at Minds # **Qualitative Research at Scale: Methods, Tools and Limits** Qualitative research at scale means getting open-ended, explain-why answers from hundreds of people or profiles instead of a dozen, then coding them into themes you can count. There are three ways to do it in 2026: AI-moderated interviews with real participants, synthetic Audiences that answer instantly, and AI-assisted coding of answers you already have. Most teams combine them._Updated October 3, 2026. Minds sells one of the approaches compared below; we say where the others are the better choice._ Qualitative research at scale means getting open-ended, explain-why answers from hundreds of people or profiles instead of a dozen, then coding them into themes you can count. In 2026 there are three practical ways to do it, and most research teams end up combining them rather than picking one. | If you need… | Use | Example tools | What you give up |
| --- | --- | --- | --- | | Real customers' words, at a few hundred interviews | AI-moderated interviews | Outset, Listen Labs, Conveo, Strella | Recruitment time and incentive cost per participant | | A fast first read on many segments or concepts before fieldwork | Synthetic Audiences | Minds | Answers are modelled, not observed, so they need validation | | Themes from answers, tickets or transcripts you already have | AI-assisted coding and repositories | Dovetail, Thematic, Minds coding | Only as good as the data you already collected | ## Why qualitative research stops scaling Classic qualitative work is limited by moderator hours, not by the number of people who could be asked. A moderator can run a handful of 45-minute interviews a day, and each transcript then has to be read and coded by hand. That is why a typical programme stops at 8 to 30 interviews. Small samples are often enough to _find_ themes. In Guest, Bunce and Johnson's widely cited 2006 study of 60 in-depth interviews, saturation occurred within the first twelve interviews, and basic elements of the main themes were present after six ([Guest et al., 2006](https://journals.sagepub.com/doi/10.1177/1525822X05279903)). Scale becomes necessary when you need to _compare_: five segments, four markets, three concepts, or the same question every month. Twelve interviews per cell turns into hundreds of interviews very quickly. ## Option 1: AI-moderated interviews with real participants AI-moderated interview platforms replace the human moderator, not the participant. You write a discussion guide, recruit or invite people, and an AI interviewer asks the questions by voice, video or text, probes vague answers and produces a summary across all sessions. Vendors in this category include Outset, Listen Labs, Conveo, Strella and GetWhy, and several of them can source participants for you._Choose this when_ the decision needs real human evidence: messaging that will carry spend, product decisions in regulated categories, or anything you will quote to leadership as "customers said". _Watch for_ the cost and time of recruitment, incentive budgets, fraud screening, and the fact that an AI moderator follows the guide you wrote, so a weak guide produces weak interviews at scale. We compare these vendors in detail in [AI customer interviews: AI-moderated tools compared](https://getminds.ai/blog/ai-customer-interviews). ## Option 2: Synthetic Audiences for a fast first read A synthetic Audience is a group of AI personas, called Minds, built to match a described population with exact quotas. In a Minds [Study](https://getminds.ai/guide/studies) you can ask that Audience open-ended questions, run in-depth interviews with individual Minds, show stimuli such as copy, landing pages, decks or concepts, and get coded themes with counts and segment comparisons in the same workflow. The same Study can also run questionnaires and deterministic methods such as MaxDiff, so qualitative and quantitative questions sit side by side._Choose this when_ you need to narrow options before paying for fieldwork: which three of twelve concepts deserve real interviews, which objections to probe, which segment reacts differently, how to word the questions. Answers arrive in minutes, and you can rerun the same questions after every change._Watch for_ the evidence boundary. Synthetic answers are modelled from public sources and the material you supply. They are directional, not a measurement of what customers will do. Minds makes the gap visible: an [Audience can be validated](https://getminds.ai/guide/audiences) against real published surveys or your own survey files, and each survey gets a score out of 100 with a 95% range. Validation needs at least 10 ready Minds and 8 fitting questions, and is included on paid plans. Our own research shows why a single good match is not proof: [a survey match does not prove user simulation](https://getminds.ai/research/survey-match-does-not-prove-user-simulation). ## Option 3: AI-assisted coding of answers you already have Many teams already hold thousands of open answers in NPS comments, support tickets, app reviews and old transcripts. Repository and analysis tools such as Dovetail and Thematic help tag and search that material, and Minds codes open Study answers against a versioned coding frame with definitions and the exact supporting passages. Minds is explicit that this is automated content analysis with model review, not human intercoder validation._Choose this when_ the evidence exists and the bottleneck is reading it. It does not answer questions nobody has asked yet. ## A workflow that combines all three 1. _Mine what you have._ Code existing open answers to list the themes and the language customers already use. 2. _Explore with a synthetic Audience._ Build an Audience for each segment in Minds, ask the open questions and show the stimuli. Compare themes by segment and cut the options down. 3. _Validate the Audience._ Run a validation against the closest published survey or your own past survey and read the score and range before you rely on the direction. 4. _Field the finalists with real people._ Take the two or three surviving options into AI-moderated interviews or a survey with recruited participants. 5. _Close the loop._ Feed the real answers back into the Minds' knowledge so the next round starts from evidence. ## Costs to plan for Minds pricing is public: a Free plan to try it, Individual at 59 EUR or 59 USD per month with 500 synthetic responses, Team at 99 EUR or 99 USD per seat per month with 4,000 synthetic responses per seat pooled across the team, and custom Enterprise volume. Individual includes 3 validations a month and Team 3 per seat. AI-moderated interview vendors mostly quote prices on request, and recruited interviews add incentives and recruitment fees per participant, so ask for a per-completed-interview price including both before you compare. ## Limits and when not to use synthetic qualitative research - Do not use synthetic answers as the final evidence for a launch, a price, a claim in advertising or anything regulated. Use real participants. - Do not use them for sensory, physical or in-store behaviour. Minds can react to descriptions and images, not taste a product. - Do not treat counts of synthetic themes as market shares. They tell you which themes are likely to matter, not how many customers hold them. - Do expect weaker results for niche populations with little public evidence. Add your own research files to the Audience, then validate. - Do keep a human researcher reviewing the coding frame and the conclusions, whichever method produced the answers. ## How to choose Pick AI-moderated interviews when you need real customers' words at volume and can wait for recruitment. Pick a synthetic Audience in Minds when you need to compare many segments or options today and will confirm the result later. Pick AI-assisted coding when you are sitting on answers nobody has read. If you are unsure, start with the cheapest step that removes the most options, and save real fieldwork for the decisions that carry money. Related reading: [what synthetic market research is](https://getminds.ai/blog/what-is-synthetic-market-research), [how to combine synthetic panels with human research](https://getminds.ai/blog/combine-synthetic-panels-with-human-research), [AI focus groups](https://getminds.ai/blog/ai-focus-group), [simulated customer interviews](https://getminds.ai/blog/simulated-customer-interviews), [how to validate synthetic panels against real data](https://getminds.ai/faq/how-to-validate-synthetic-panels-against-real-data), and [synthetic research vs traditional market research](https://getminds.ai/comparison/synthetic-research-vs-traditional-market-research). ## Sources - Guest, G., Bunce, A. and Johnson, L. (2006). [How Many Interviews Are Enough?](https://journals.sagepub.com/doi/10.1177/1525822X05279903) _Field Methods_, 18(1). - Minds plan prices and allowances: [getminds.ai/pricing](https://getminds.ai/pricing), as of October 2026. - Minds research: [A survey match does not prove user simulation](https://getminds.ai/research/survey-match-does-not-prove-user-simulation). ## **Frequently asked questions**### **What does qualitative research at scale mean?** It means collecting open-ended answers, the reasons behind a choice rather than only the choice, from hundreds of participants or profiles instead of the 8 to 30 a classic interview or focus group programme reaches, and then coding those answers into themes you can count and compare between segments. ### **Can AI moderate real customer interviews?** Yes. AI-moderated interview platforms such as Outset, Listen Labs and Conveo send a link to real participants, ask the discussion guide by voice or text, follow up on vague answers and summarise the results. You still recruit the people, and the answers are real human evidence. ### **Is synthetic qualitative research the same as interviewing customers?** No. A synthetic Audience in Minds answers in seconds and costs a fraction of fieldwork, but its answers are modelled, not observed. Use it to explore and narrow options, check it against real survey data with a validation score, and confirm high-stakes decisions with real people. ### **How many interviews do you need before themes stop changing?** It depends on how varied the population is. In a frequently cited study of 60 in-depth interviews, Guest, Bunce and Johnson found that saturation occurred within the first twelve interviews, which is why scale matters most when you need to compare several segments rather than describe one. ### **What does Minds cost for qualitative work?** Minds has a Free plan, Individual at 59 EUR or 59 USD per month with 500 synthetic responses, and Team at 99 EUR or 99 USD per seat per month with 4,000 synthetic responses per seat pooled across the team. Enterprise volume is custom. [Minds](https://getminds.ai/)© 2026 Minds. Your target audience. AI-driven and grounded in transparent evidence. Build within minutes. [Minds on X (Twitter)](https://x.com/mindsai_co) [Minds on LinkedIn](https://www.linkedin.com/company/mindsaicompany/) [Minds on Instagram](https://www.instagram.com/getminds.ai/)Minds is part of [![ESOMAR](https://getminds.ai/images/newsroom/logos/esomar-logo.svg)ESOMAR](https://esomar.org/) [![GreenBook](https://getminds.ai/images/newsroom/logos/greenbook.svg)GreenBook Directory](https://greenbook.org/company/Minds) [![Insight Platforms](https://getminds.ai/images/newsroom/logos/insight-platforms.png)Insight Platforms](https://www.insightplatforms.com/platforms/minds/) [![Capterra](https://getminds.ai/images/newsroom/logos/capterra.svg)Capterra](https://www.capterra.com/p/10046203/Minds/) [![G2](https://getminds.ai/images/newsroom/logos/g2.svg)G2](https://www.g2.com/products/minds/reviews) [![CSSDA Best UX Design Award](https://getminds.ai/images/newsroom/logos/cssda-best-ux-award.png)CSSDA Best UX Design Award](https://www.cssdesignawards.com/) [![CSSDA Best Innovation Award](https://getminds.ai/images/newsroom/logos/cssda-best-innovation-award.png)CSSDA Best Innovation Award](https://www.cssdesignawards.com/) [![CSSDA Best UI Design Award](https://getminds.ai/images/newsroom/logos/cssda-best-ui-award.png)CSSDA Best UI Design Award](https://www.cssdesignawards.com/)