Designing a VoC research plan
- Define the research questions:
- What do you need to learn from participants? Be specific: not "what do customers want?" but "what are the three most frustrating moments in accessing this service?" or "what would participants need in order to recommend this program to someone else?"
- Distinguish between: stated needs (what people say they want), latent needs (what they want but don't articulate), and excited needs (features they didn't know they wanted but would love).
- Select research methods:
VoC typically combines multiple data collection methods:
- In-depth interviews (IDIs): One-on-one conversations, 45–90 minutes, following a semi-structured guide. Best for exploring complex experiences, sensitive topics, and minority voices that might be suppressed in groups.
- Focus groups: Moderated group discussions, 6–10 participants, 90–120 minutes. Best for exploring shared experiences, social norms, and how ideas develop in conversation.
- Surveys: Quantitative reach to larger populations. Best for validating qualitative findings and measuring prevalence of specific needs or experiences. Weak at surfacing unexpected insights.
- Ethnographic observation: Observing participants in their natural context — how they actually use a service, not how they describe using it. Powerful for surfacing workarounds and unspoken practices.
- Digital feedback channels: Online reviews, social media, customer service transcripts, app analytics. Rich source of unsolicited, authentic voice — often overlooked as a VoC input.
- Recruit participants:
- Define participant profiles based on who uses or is affected by the service. Deliberately include: current heavy users, occasional users, lapsed users (who stopped), and people who never engaged but might have.
- Avoid recruiting only satisfied customers or organizational contacts — the most important insights often come from people at the margins of the system.
Step-by-step instructions
1
Open with context and consent (5 min)
- Explain the purpose of the research (not to evaluate the participant — to understand their experience).
- Confirm that responses are confidential and will be anonymized in reporting.
- Obtain informed consent for recording if applicable.
- Clarify that there are no right or wrong answers — honest experience is the goal.
2
Warm-up: establish context (10 min)
- Ask about the participant's relationship with the service or program: How long have they been using it? How did they first encounter it? What do they use it for?
- Goal: establish rapport and ground the conversation in specific, real experience rather than general opinion.
3
Core questions: explore experience in depth (30–50 min)
- Use open-ended questions that invite narrative: "Tell me about the last time you used [service]..." "Walk me through what happens when you [task]..." "Describe a moment when this worked really well for you..." "Tell me about a time when it didn't work the way you needed..."
- Follow up on everything emotional: when participants express frustration, surprise, or satisfaction, probe further. "You mentioned that was frustrating — can you tell me more about that?"
- Ask about workarounds: "When X doesn't work the way you need, what do you do instead?" Workarounds reveal unmet needs that participants have already half-solved.
- Ask about outcomes: "When this works well, what does that make possible for you?" Understanding the downstream benefit reveals what the service is actually for, from the user's perspective.
4
Future needs (10 min)
- "If you could change one thing about this [service/program/product], what would it be?"
- "Imagine it's [year], and this is exactly what it should be. What does that look like?"
- "Who else should we talk to — someone whose experience we might be missing?"
5
Close and debrief (5 min)
- Summarize what you heard and check: "Is there anything important I missed or misunderstood?"
- Thank participant and explain what will happen with their input.
- Write field notes immediately after the interview while impressions are fresh — not just what was said, but how: tone, hesitation, body language.
Analysis: from raw voice to actionable insight
- Transcribe and code: Transcribe all interviews. Code responses by theme: what topics come up repeatedly? What are the emotional peaks (positive and negative)? What unmet needs are expressed?
- Affinity mapping: Group coded insights into clusters. Common clusters emerge: "accessing the service", "receiving the service", "follow-up and support", "barriers and frustrations", "things that work well."
- Prioritize by frequency and intensity: An insight mentioned by 15 of 20 participants with strong emotion is more action-worthy than one mentioned once. But rare insights from severely underserved users may still be high priority.
- Translate to needs statements: Convert raw voice into needs statements in the format: "[Participant type] needs to [do something] so that [outcome]." Example: "A first-time user needs to understand what documents to bring before they arrive so that they don't have to make a second trip."
- Report and close the loop: Share findings with the team and decision-makers. And critically: tell participants what you learned and what you will do with it. VoC without this feedback loop is extractive and undermines trust in future research.
Materials
Interview guide (semi-structured question list, not a script)
Recording device and consent forms (if recording)
Transcription tool or service
Analysis workspace (physical or digital affinity mapping)
Participant recruitment list and screening criteria
Incentives for participation (gift cards, acknowledgment, time compensation)
Practical recommendations
Listen for what is not said. The pauses, the topics participants avoid, the workarounds they describe without labeling them — these are often the most valuable signals. Train interviewers to notice and probe silence and avoidance.
Separate data collection from analysis. Do not analyze while collecting. Analysis in the field introduces bias — the interviewer starts confirming patterns they've spotted rather than staying open to surprise.
Recruit beyond the satisfied majority. Customer satisfaction surveys systematically miss the most important voices: people who stopped using the service, people who never started, and people who use it but have adapted their behavior to work around its failures.
Bring internal stakeholders into the analysis. When the team that will act on findings participates in the analysis (listening to interview recordings, doing affinity mapping), the insights land differently than when they receive a report. The analysis process builds empathy that no document can replicate.
Set expectations about what will and won't change. Participants who share honest feedback and never hear what happened to it become less likely to participate in future research. Even if most findings can't be acted on immediately, communicate what was heard and what the constraints are.
Inspiration
Application examples by context:
- Healthcare: A hospital conducts VoC interviews with patients discharged in the past 90 days and discovers that the most common complaint — long wait times — masks a deeper problem: patients cannot accurately predict how long they will wait, which makes waiting feel longer and more distressing. The solution is not speed but communication.
- Education: A university runs VoC with first-year students who dropped out and discovers a consistent pattern: the decision to leave was made in the third week, triggered by a combination of academic isolation and administrative confusion — two problems that orientation programs don't address.
- Social services: An NGO conducts VoC with beneficiaries of a food assistance program and learns that the most significant barrier is not food availability but the shame participants feel in the application process — leading to a redesign of the intake experience.
- Public policy: A government agency conducts VoC before redesigning a benefits application process and discovers that 60% of eligible non-applicants are excluded not by language or documentation but by appointment scheduling — they can't take time off work for a mid-day in-person appointment.
- Six Sigma / Lean: VoC is a core input to the DMAIC (Define-Measure-Analyze-Improve-Control) cycle. Critical to Quality (CTQ) requirements are derived directly from VoC data.
- Design Thinking: The Empathize phase is VoC methodology applied to innovation: in-depth interviews and observation before any solution is designed.
- Net Promoter System: The NPS question ("How likely are you to recommend us?") is a single-question VoC instrument. Its value is in the follow-up "why" question, which generates qualitative VoC data at scale.
Consulted sources
- Griffin, A. & Hauser, J. R. (1993). The voice of the customer. Marketing Science, 12(1), 1–27.
- Ulwick, A. W. (2002). Turn customer input into innovation. Harvard Business Review, 80(1), 91–97.
- Patton, M. Q. (2014). Qualitative Research and Evaluation Methods (4th ed.). SAGE Publications.
- Burchill, G. & Brodie, C. H. (1997). Voices Into Choices: Acting on the Voice of the Customer. Joiner Publications.
