In-Depth Interview Method: How to Plan, Size and Brief One

The method answers "why" and "how". It does not answer "how many". Nine to seventeen interviews cannot give you a percentage, so do not report one.

Plan 45 to 60 minutes per interview (our recommendation). Go shorter for one narrow topic and longer only for step-by-step journeys.

Our recommendation for choosing a method:

If your question has a "why" part and a "how many" part, run interviews first to find the reasons, then a survey to size them.

Plan 9 to 17 interviews for each distinct audience when the group is fairly uniform and the question is narrow. A 2022 systematic review of empirical saturation tests found saturation, the point where new interviews stop adding new themes, within that range for relatively homogeneous populations with narrow aims.

A 2006 analysis of 60 interviews with women in two West African countries (Guest, Bunce and Johnson) found saturation by 12 interviews, with the main themes visible by the sixth. Both results hold for narrow questions and similar participants.

The 2024 analysis cited in the introduction reached 15 to 23 interviews for near saturation and 30 to 67 for true saturation. It found that less structured guides, heavily inductive analysis (themes built from the data rather than from a fixed list) and participants less knowledgeable about the topic can push the number up. Our reading: treat 9 to 17 as the lower bound for a uniform audience, a narrow question and a structured guide. Loosen the guide, widen the question, mix audiences or analyse inductively, and plan toward the 15 to 23 range and, for some designs, 30 and more.

The size of the study grows with the number of segments, meaning groups you want to compare, such as users in Delhi and users in Indore, or current and lapsed customers. Each segment needs its own set.

Our worked model: interviews = segments × 9 to 17. Interviewing hours = interviews × 45 to 60 minutes (9 × 45 minutes = 6.75 hours; 17 × 60 minutes = 17 hours). Recruited = interviews + 20%, rounded up per segment (9 becomes 11; 17 becomes 21). The 20% is our recommendation for no-shows; adjust it after your first pilot.

Hours cover the sessions only. They exclude notes and analysis.

Our stopping rule: book six per segment, then add three at a time. Stop when the last three interviews added no theme missing from your notes. Pass: zero new themes in the last three. Fail: any new theme, so add three more. Our sample size guide goes deeper on this.

Six question habits weaken IDI data, according to Nielsen Norman Group: opening with screener questions, asking only about typical behaviour, hypotheticals, leading clarifications, compound questions and ambiguous questions. A screener is a short questionnaire that checks whether someone qualifies for the study. Each habit has a fix: ask about one specific past event, one question at a time.

Goals that are too broad, such as "learn about users", also produce interviews with nothing actionable. Write the one decision the research feeds before writing a single question.

For most commercial studies we recommend a semi-structured guide: fixed core questions with room to follow the participant. Answers stay comparable across people. Fill in this skeleton:

After a short answer, wait a few seconds before moving on. The fuller answer often comes in the pause. Debrief the same day, in three lines per interview: what surprised you, what contradicted your assumption, what to probe next time.

A recruitment brief should fix eight things before anyone is invited: the decision, the segments, the interviews per segment, session length, mode and language, consent and recording wording, incentive terms, and deliverables. Each has a test you can run.

This is how we work at Qualfacto, a qualitative research participant panel in India operated by Think Design Research Services. You share your brief with us. We review the objective, method, audience, geography, sample, timelines and recruitment criteria, then assess feasibility and send a tailored proposal and quote. We have no fixed rate card because cost depends on the method, audience, recruitment difficulty and scope. After you agree the scope, we recruit and screen participants and schedule the sessions. We process participant incentives after validation, and you receive the agreed deliverables. Our recruitment covers metros, Tier 1 cities (India's largest cities), Tier 2 cities (mid-sized cities such as Indore or Coimbatore), and semi-urban and rural locations, with multilingual fieldwork assessed project by project. For a first project, we can offer a small pilot or introductory arrangement so you can judge participant quality before committing. For more on this stage, read how to recruit participants for a study in India.

To start, share your brief with Qualfacto and we will tell you what is feasible.

Tell every participant, before the interview starts, what the research is for, what you will record, who will see it and that they can stop at any time, then record their agreement. India's Digital Personal Data Protection Act, 2023 governs personal data, and on our reading an interview recording of an identifiable person is personal data.

The Digital Personal Data Protection Rules were notified in November 2025. As of 10 October 2026, the notified deadline for the main duties on organisations that collect data, including consent notices and security safeguards, is mid-May 2027, 18 months after notification. MeitY, the Ministry of Electronics and Information Technology, consulted in January 2026 on shortening that to 12 months, which would be mid-November 2026, but no amending notification has been published. Re-check before you rely on either date.

Section 17(2)(b) of the Act exempts research processing that is not used for decisions about an individual and that follows the standards in the Rules' Second Schedule. Check with counsel whether your study qualifies, because commercial studies often involve follow-up contact or incentives. Do not wait for the deadline to get consent right. The ICC/ESOMAR Code (International Chamber of Commerce and ESOMAR, the global body for market, opinion and social research) already sets professional standards for research.