How it works
How to run AI interviews: from goal to insight in 3 steps
Running AI interviews with BlueLandscape takes three steps: describe your goal, share a chat link, and read the insights it pulls out for you. No research background needed. You write the brief in plain language, and the AI does the interviewing, the follow-ups, and the analysis.
A static survey forces everyone down the same fixed path. An AI interview follows the conversation instead — it reads each answer, asks the obvious next question, and digs into the parts a form would skip right past. That is the whole difference, and it is why the setup below stays so short.
Step 01
Describe your research goal
Tell the AI what you want to learn — in plain language. It designs the interview flow, picks the right approach (discovery, validation, retrospective), and sets the tone for your audience. There are no question banks to fill and no branching logic to wire up. Want to add a constraint, like "keep it under five minutes"? Just say so, and the brief updates.
Step 02
Share the link with respondents
You get one short chat link (/chat/<id>). Drop it in an email, a Slack message, or a QR code. Each respondent opens it on their own time, on any device — async, no calendar invite, no call to schedule. The AI interviews them one-on-one: it paraphrases to check it understood, asks "could you say more about that?" when an answer is thin, and keeps the conversation on track. It speaks the respondent's own language automatically.
Step 03
Get structured insights
Every conversation gets its own AI summary, and themes are extracted across all of them — so the patterns that show up in respondent after respondent surface without you tagging a single quote. No transcription, no coding framework, no spreadsheet of open text. You read findings you can act on, and you can generate a full report over the results when you are ready to share.
What the AI does during each interview
Behind those three steps, a fair amount happens on its own. Here is what every session includes:
- Adaptive probing — the AI follows up on interesting answers and pushes past vague ones, the way a good interviewer would.
- Async chat link — respondents answer whenever suits them; you never coordinate a calendar.
- Per-session summaries — each conversation is condensed into the points that matter.
- Auto-extracted themes — cross-respondent patterns are grouped for you, no manual tagging.
- PII redaction — IBAN, credit card, BSN, phone, and email are stripped at intake unless you opt a question into an allowlist.
- EU data residency — everything stays in Amsterdam, GDPR-compliant, consent and timestamp logged.
A quick example
Imagine you are a three-person SaaS team about to raise a plan from €19 to €29 a month. You write one line — "find out how our paying users feel about a price increase and what would make it fair" — and send the link to forty customers. Over the next few days they reply between meetings. One mentions they would stay "if onboarding got faster," so the AI asks what slowed them down. By Friday you open the results and see "onboarding friction" sitting at the top of the themes, with the exact quotes underneath. You never wrote a follow-up question or read a transcript.
Common questions
Do I need to write the questions myself?
No. You describe the goal and the AI writes and orders the questions, then adapts them live for each person. There is no question tree to maintain.
How long does setup take?
Minutes. The longest part is deciding what you actually want to learn — once that is clear, the brief is a sentence or two.
What does it cost to run?
Pay-per-use: roughly €0.05 to €0.50 per completed session depending on length, with no per-seat fees and a €50 monthly cap by default. See pricing for the detail.
Is respondent data safe?
Personal data is auto-redacted at intake, everything stays in the EU, and it is GDPR-compliant with consent logged.
Learn more: What are AI-moderated interviews? · Customer discovery · Employee feedback · Pricing
Updated June 2026.