"Depth interview," "user interview," "focus group" — these terms get used interchangeably in practice, even though they're three different research methods. So what's the actual difference, and when do you use which?
This article lays out what a depth interview is in its proper sense, how it differs from adjacent methods, the disciplines that determine whether the depth actually surfaces, and where it fits in a marketing research program.
Definition
A depth interview (also "in-depth interview" or IDI) is:
- A 1-on-1 qualitative research method aimed at deep motivation
- 60–120 minutes per session
- Researcher (moderator) and one respondent
- Semi-structured to unstructured conversation
- Designed to surface motivations, values, and unconscious patterns the respondent has not articulated before
As the "depth" in the name suggests, the goal is to dig into motivations that don't surface in shallower formats.
How it differs from adjacent methods
vs. user interviews
| Dimension | Depth interview | User interview |
|---|---|---|
| Focus | Motivation, the unconscious | Actual usage, behavior |
| Question style | Abstract, hypothetical | Concrete, behavior-grounded |
| Used for | Marketing strategy | Product improvement |
| Length | 60–120 min | 30–60 min |
User interviews center on "how do you use the product?" Depth interviews center on "what are you actually looking for? what are you afraid of?"
In practice, teams often combine them, or run one as primary with elements of the other woven in.
vs. focus groups (FGI)
| Dimension | Depth interview | Focus group |
|---|---|---|
| Participants | 1 | 6–8 |
| Format | 1-on-1 | Group discussion |
| Strength | Individual depth | Interaction, multiple perspectives |
| Weakness | Per-session cost | Conformity pressure suppresses honesty |
| Best for | Sensitive, personal topics | Topics meant to be debated |
Sensitive or personal topics belong in depth interviews. Topics where you want a range of perspectives belong in focus groups.
vs. ethnography
| Dimension | Depth interview | Ethnography |
|---|---|---|
| Primary method | Conversation | Observation |
| What you capture | What people articulate | Behavior and environment |
| Where | Conference room, online | Their home, their workplace |
| Effort | Medium | Large |
Ethnography surfaces what people don't put into words. Depth interviews surface what they can put into words once you create the space.
How to run one
Step 1: Selection
n=5–15 is standard. Depth interviews tend to prefer "distinctive individuals" over "typical customers":
- High-loyalty customers
- Customers using the product in unusual ways
- Customers right before or after churn
- Switchers from a competitor
- Out-of-target customers who use the product anyway
Step 2: Script
0–10 min: Rapport
- Introductions
- Goal, runtime, recording consent
- Light, relaxing conversation
10–30 min: Background
- Lifestyle, values, what they care about
- Their general relationship to the category
30–70 min: The deep block (most important)
- Concrete stories
- Emotional movement
- Tensions and contradictions
- Hypothetical scenarios ("if you couldn't...")
70–90 min: Hypothesis testing
- Float your pre-built hypotheses
- Observe reaction
- Leave room for "anything else"
90–120 min: Close
- "Anything you didn't say?"
- Honorarium logistics
- Permission for follow-up
Step 3: In the room
- Recording is non-negotiable (with consent)
- Notes: keep them short
- Watch facial expressions, tone, silences
- Don't lead, don't rush, don't decide for them
Step 4: Analysis
1. Transcribe everything
2. Pull striking quotes verbatim (don't paraphrase)
3. Cluster by theme
4. Find shared patterns and outliers
5. Articulate the insight hypothesis
Step 5: Reporting
- Report in stories, not numbers
- Quote respondents verbatim
- Separate "hypothesis" from "validated fact"
- Always include a next action
The hard part — what makes depth "depth"
"We did an interview" and "we did a depth interview" are different things. The difference is whether depth actually showed up.
Signs of a shallow interview
- You're working through your prepared questions in order
- You take the respondent's words at face value
- Sessions end on words like "convenient" or "satisfied"
- You walk out with no new finding
Signs of a deep interview
- Their answers pushed you in unexpected directions
- You found contradictions and dug into them
- There's a moment when the respondent says "huh, I hadn't thought about that before"
- You walk out with a clear new hypothesis
Ten prompts that surface depth
1. "Can you give me a specific situation?" (concretize)
2. "How did that feel at the time?" (emotion)
3. "What would you have done if X?" (hypothetical)
4. "Did you consider other options?" (comparison)
5. "How would you explain it to someone else?" (third-party lens)
6. "Tell me a little more about that" (probe)
7. "What part did you actually dislike?" (flip)
8. "Why do you think you felt that way?" (reason)
9. "Have you had similar experiences elsewhere?" (pattern)
10. "If you could rewind, what would you do?" (hypothetical)
Common failure modes
Failure 1: Treating the script as a checklist
The script gets mechanically worked through and the deep probes never happen.
Fix: the script is a drawer, not a checklist. Follow the conversation.
Failure 2: Talking too much
The researcher shares opinions and hypotheses. The respondent shifts into "agree with the smart person" mode and stops disclosing.
Fix: target ratio is researcher 20% / respondent 80%.
Failure 3: Filling silences
When the respondent goes quiet, the moderator panics and asks the next question — losing the deep response that was about to come out.
Fix: 5 seconds of silence is fine.
Failure 4: Letting unclear statements pass
When you don't quite understand what they meant, you pretend you did and move on, losing the deepest probe of the session.
Fix: always check — "when you said 'X' just now, can you tell me a little more?"
Failure 5: Treating n=5 hypotheses as facts
The hypotheses from 5 depth interviews get circulated as "what customers think" without quantitative validation.
Fix: distinguish hypothesis vs. validated explicitly. Run a survey at scale before calling it a fact.
Where depth interviews fit in a research program
Strategy phase
New market entry, brand repositioning, new business creation — the hypothesis-generation phase. n=10–15 to build deep customer understanding.
Improvement phase
Product / UX / CX work — combined with user interviews. n=5–10 to surface concrete improvement opportunities.
Validation phase
Hypothesis validation is usually better done with surveys at scale. Depth interviews are rarely the right tool here.
Steady state
For mature products, a quarterly cadence of 1–2 depth interviews keeps customer understanding current.
Cost and effort
Self-run
- Recruiting: 10–20 hours per project
- Sessions: 3–4 hours per respondent × 5–15 respondents = 15–60 hours
- Transcription and analysis: 5–10 hours per respondent
- Report: 10–20 hours
- Total: 50–200 hours per project
Outsourced
- A typical project: low five figures to low six figures USD (recruiting + sessions + analysis)
- Per-respondent vendor rates: hundreds of USD
- Use when speed matters or in-house capability is missing
Where Repoan fits
Repoan complements depth interviews on either side of the session:
- Screening surveys — recruit the right respondents
- Post-interview scale validation — convert n=5 hypotheses into questions to run across all customers
- Continuous observation — monthly surveys that signal when the next depth interview is needed
Summary
Depth interviews:
- 1-on-1 qualitative research for deep motivation (60–120 min)
- More abstract and hypothetical than user interviews
- Surface more honesty than focus groups
- Combine with ethnography for behavior + consciousness
- Best in the hypothesis-generation phase
- Always validate at scale with surveys before acting
The depth comes from interview craft and the researcher's stance. As marketing leans further on AI, the relative value of depth interviews — the part AI can't run for you — is going up, not down.