AI-Augmented Product Discovery and User Research

Here are transcripts from 8 user interviews about our
accounts payable workflow: [paste transcripts]

Synthesize these interviews:
1. The top 3 pain points mentioned by more than half of users
2. The top 3 pain points mentioned by only some users —
   describe which user types experience each
3. Workarounds users have developed for existing problems
   (signs of unmet needs)
4. Statements that suggest an emotional response
   (frustration, anxiety, pride) — these indicate high-value problems
5. Feature requests — and for each, what underlying need
   it suggests (the request is the symptom; the need is the cause)
6. Surprising findings that contradict common assumptions

Format: numbered list per category. Quote the user (with ID,
not name) to support each finding.

Here is the survey data from 214 responses about feature priority: [paste]

Analyze:
1. Top 5 features by overall priority score
2. How priority differs by user segment (if segmentation data provided)
3. Features where there is high disagreement between respondents
   (some rate very high, some rate very low) — these indicate
   the feature has strong advocates and detractors
4. Open-text responses: common themes, notable outliers, verbatim
   quotes that illustrate the themes clearly
5. Gaps: problems mentioned in open text that no listed feature addresses

Based on this research summary about our AP users: [paste]

Generate product hypotheses in this format:
"We believe [user type] experiences [problem] when [situation].
We think [proposed feature/change] will [expected outcome].
We will know this is true when [measurable signal]."

Generate 8 hypotheses — mix of high-confidence (well-supported
by research) and low-confidence (interesting but not yet validated).
Flag each as HIGH/MEDIUM/LOW confidence based on research support.

I'm doing competitive research for a B2B accounts payable
automation product. We're targeting mid-size manufacturing companies.

For each of these competitors: [list competitor names]

Analyze based on publicly available information:
1. Target customer and positioning
2. Core feature set
3. Stated differentiation
4. Pricing model and entry point
5. Customer reviews (common praise and common complaints)
6. Recent product announcements (what direction are they going?)

Then:
- What problems do all competitors address (table stakes)?
- What problems does no competitor address well (opportunity gaps)?
- What differentiation claims seem validated vs. marketing copy?

Based on these research insights: [paste synthesis notes]

Develop 3 user personas for the AP automation product.

For each persona:
- Name and role (fictional but representative)
- Day-in-the-life description (what their work actually looks like)
- Primary goals (what they're trying to accomplish)
- Key pain points (what makes their job harder)
- Technical comfort level
- Decision-making role (do they buy, use, or influence the product?)
- What they need from a solution (not features — outcomes)
- What would make them resist adopting a new system

Ground each persona in specific quotes from the research.
Flag any characteristics that are assumed vs. research-supported.

Based on this user research: [paste]

Write a problem statement for the AP automation product opportunity
using this format:

[Target user] experiences [problem] when [context].
This causes [impact on their work/business].
Today they solve this by [current workaround].
A better solution would allow them to [desired outcome]
without [trade-off they're currently forced to make].

Then validate this problem statement:
1. Is it solution-neutral (doesn't prescribe an answer)?
2. Is it specific enough to be testable?
3. Does it reflect the research, or does it reflect
   what the business wants the problem to be?