Key Informant Interviews (KII): Meaning, Method and When to Use Them

When are key informant interviews the right method?

Key informant interviews (KIIs) are the right method when you need to understand context, mechanisms, perceptions, or decision-making processes that structured questionnaires cannot capture. Use them when:

• Explanatory depth is needed: survey data shows what happened; KIIs explain why
• Specific knowledge holders must be reached: government officials, program managers, community leaders, technical specialists, or service providers who hold unique perspectives
• Triangulation is required: KIIs complement quantitative data by providing explanatory context and validating or challenging survey findings
• Rapid assessment timelines: KIIs can be designed and conducted in days, making them well-suited to rapid assessments, situational analyses, and emergency contexts

KIIs are not appropriate for collecting data that should be statistically representative (use surveys), for group dynamics or community-level perspectives (use focus group discussions), or for direct observation of behavior (use observation methods).

How It Works

Step 1: Define the purpose and key informant types

What specific information do you need, and who holds it? Identify the types of informants: program staff (implementation perspective), community leaders (community perspective), service providers (delivery perspective), government officials (policy perspective), beneficiaries (experience perspective). Aim for diversity across perspective types.

Step 2: Develop the interview guide

Write a semi-structured guide with 5-10 open-ended question areas (not scripts). Organize from general to specific. Ensure each question area maps to an evaluation question. Include probes for follow-up. The guide should be a reference, not a rigid script - follow where the interview leads.

Step 3: Select informants purposively

Purposive sampling is standard for KIIs: select individuals because of their specific knowledge or position, not randomly. Aim for 20-40 informants for most evaluations, targeting saturation - when new interviews are producing no new themes.

Step 4: Conduct the interviews

Use a two-person team where possible: one interviewer, one notetaker. Record with consent. Build rapport before sensitive questions. Use active listening, silence, and probes to draw out depth.

Step 5: Analyze thematically

Organize notes or transcripts by theme, not by interview. Look for patterns, contradictions, and outliers. Compare findings across informant types.

Step 6: Triangulate with other data sources

KII findings should be triangulated with survey data, document review, and (where applicable) focus group discussions before conclusions are drawn. No single KII finding should drive a major conclusion without corroboration.

Key Components

• Interview guide: 5-10 open-ended question areas with probes, organized by theme
• Informant typology: a defined set of informant roles to be covered, with rationale
• Sampling matrix: tracking which informant types have been interviewed and ensuring diversity
• Note-taking protocol: standardised approach to capturing verbatim quotes and observations
• Analysis framework: a thematic coding structure aligned to evaluation questions
• Triangulation matrix: tracking how KII findings compare to other data sources

Best Practices

Use semi-structured, not structured. Fully scripted interviews produce survey-quality data through an inefficient method. The value of KIIs comes from following unexpected threads. Keep the guide as a reference, not a constraint.

Prioritize listening over talking. Interviewers who fill silences with more questions get less data. Silence prompts elaboration. Restate, don't redirect.

Record with consent, transcribe selectively. Full transcription of all interviews is expensive and rarely necessary. Transcribe verbatim only key passages; use detailed notes for the rest.

Include data quality checks. If a finding appears in only one interview and contradicts all other data, treat it with caution. Note it as an outlier, not a conclusion.

Debrief after each interview. Immediately after each interview, the team should spend 10 minutes noting key findings, surprises, and follow-up questions. This debriefing captures context that does not appear in notes and guides the next interview.

Common Mistakes

Over-relying on official informants. Program managers and government officials are easy to reach but may give politically safe rather than candid responses. Balance official perspectives with community-level informants.

Leading questions. "Would you agree that the program improved access?" is a leading question. "How has access changed since the program started?" is not. Poor question design in KIIs produces confirmation bias, not evidence.

No analysis framework. Analyzing KII data without a structure ("just reading the notes") produces vague, anecdotal findings. Use a thematic coding framework aligned to evaluation questions.

Treating KII as substitute for survey. KIIs provide depth, not representation. Findings from 25 interviews cannot be extrapolated to a population of 10,000. Present findings as perspectives, not statistics.

Examples

Mid-term review, West Africa. A donor-funded governance program in Ghana conducted 32 KIIs as part of its mid-term review, covering district officials, CSO representatives, community leaders, and program staff. The interviews revealed a mechanism not visible in monitoring data: the program's most effective activities were those where district officials participated as co-facilitators rather than passive recipients. This finding informed a design revision for Phase 2.

Rapid assessment, East Africa. Following a drought in the Horn of Africa, an OCHA-funded team conducted 15 KIIs with local government, NGO staff, and community leaders over 3 days to assess the most critical gaps in the humanitarian response. KII findings were triangulated with observation data and secondary data from satellite imagery analysis. The combined assessment informed resource prioritisation for the following 30 days.