These prompts support the four most time-consuming steps in an Outcome-Driven Innovation (ODI) project: extracting desired outcome statements from qualitative interviews, consolidating duplicates across interviews, analyzing quantitative survey data, and profiling outcome-based segments. Each prompt is designed to work with any LLM and follows the ODI methodology developed by Tony Ulwick.
Prompt 1: Extract outcome statements from interview transcripts
You are an ODI (Outcome-Driven Innovation) researcher trained in Tony Ulwick's methodology.
I will provide a transcript of an interview with [role/type of participant] about the job of [job-to-be-done definition].
Your task:
1. Read the transcript and identify every moment where the participant describes what success looks like, what could go wrong, what they measure or care about when performing any step of this job.
2. Convert each identified moment into a formal desired outcome statement using this format:
[Direction] the [unit of measure] it takes to [object of control] [contextual clarifier]
Directions: "minimize", "increase", "reduce the likelihood of"
3. Group the outcomes by job step (define, locate, prepare, confirm, execute, monitor, modify, conclude).
4. For each outcome, cite the exact quote from the transcript that supports it.
Rules:
- Every outcome must be measurable, solution-independent, and stable over time.
- Do not include feature requests, solutions, or product references.
- Do not use vague words like "easily", "better", "quickly" — be specific about what is being measured.
- Aim for 15-25 outcomes per interview.
Transcript:
[paste transcript here]Prompt 2: Consolidate and de-duplicate outcome statements
You are consolidating outcome statements from [N] ODI interviews about the job of [job-to-be-done definition].
Below is the full list of [total number] raw outcome statements, grouped by interview.
Your task:
1. Identify outcomes that express the same underlying need in different words. Group them into clusters.
2. For each cluster, propose a single consolidated outcome statement that captures the core need in standard ODI format: [direction] the [metric] it takes to [object] [clarifier].
3. Flag any outcomes that violate ODI rules (solution-dependent, not measurable, vague, or unstable over time) and suggest fixes.
4. Organize the final list by job map step.
5. Produce a summary: total raw outcomes, total consolidated outcomes, outcomes per job step.
Target: 100-150 consolidated outcomes covering all steps of the job map.
Raw outcomes:
[paste all outcomes here]Prompt 3: Analyze opportunity scores and identify top opportunities
I have completed an ODI survey for the job of [job-to-be-done definition] with [N] respondents rating [M] outcomes on importance (1-5) and satisfaction (1-5).
The data is in [format: CSV/spreadsheet] with columns: outcome_id, outcome_text, job_step, mean_importance, mean_satisfaction.
Your task:
1. Calculate the opportunity score for each outcome: opportunity = importance + max(0, importance - satisfaction). Use the mean scores.
2. Rank all outcomes by opportunity score (highest first).
3. Identify the top 15 underserved outcomes (opportunity > 10) and explain why each represents an innovation opportunity.
4. Identify overserved outcomes (opportunity < 7, importance < 3) — these are candidates for cost reduction or feature removal.
5. Create a narrative summary for a product team: where is the market underserved? What themes emerge? What do the top opportunities have in common?
6. Suggest how to visualize this as an opportunity landscape chart.
Data:
[paste or attach data]Prompt 4: Generate outcome-based segment profiles
I ran cluster analysis on ODI survey data for the job of [job-to-be-done definition]. The analysis identified [N] segments.
For each segment, I have:
- Segment size (% of respondents)
- Top 10 underserved outcomes (with opportunity scores)
- Top 5 overserved outcomes
- Demographic/behavioral profile data
Your task:
1. Name each segment with a descriptive label that captures its defining unmet need (not demographics).
2. Write a 2-3 paragraph profile for each segment: who they are, what they struggle with, what they do not care about, and what kind of product would win them.
3. Compare segments: which outcomes are shared across segments (table stakes) and which are unique to specific segments (differentiation opportunities)?
4. Recommend an innovation strategy for each segment: differentiated (many underserved outcomes), disruptive (overserved segment willing to trade features for simplicity/price), or dominant (both under and overserved outcomes).
5. Identify which segments represent the highest-value targets and why.
Segment data:
[paste segment data here]