Competencies
The interviewing standards behind Juno conversations - Set and reviewed by qualitative researchers with 25+ years’ experience.
Research discipline
Disciplined questioning
Juno is built around the disciplines a trained researcher brings to an interview: questions earn their place, answers are weighed as evidence, and themes are pressure-tested rather than simply collected.
Brief-aligned questioning
Before it is asked, each question is checked against the decision the research has to inform, keeping the conversation tied to the brief.
Evidence-weighted answers
Concrete, spontaneous, lived detail is told apart from rehearsed generalities — so what surfaces is what people actually do, not what sounds good to say.
Active disconfirmation
Juno does not just gather support for the likely answer. It probes boundaries, seeks exceptions, and tests what would prove a theme wrong.
Mechanism over anecdote
A one-off story is separated from a repeatable pattern by surfacing the conditions underneath it — what happened once, and what is likely to happen again.
Conversation
A conversation, not a script
Paced, responsive and unscripted — the conversation follows the person while the research keeps its shape.
Breadth-first coverage
Juno maps the decision landscape before going deep on a single thread, reducing the chance that an early tangent crowds out ground that still matters.
Adaptive pacing
Juno adjusts to the time remaining and how engaged the participant is — going deeper while a thread is earning something, and moving on when it is not.
Natural flow
Brief, neutral acknowledgements and unscripted phrasing keep the conversation natural — without warmth that rewards a particular answer.
Sensitive-topic handling
On difficult subjects — money, health, family — Juno signals where a question is going, offers a way past it, and proceeds with care rather than pressing. The full duty-of-care standard lives under Safety.
No mechanical repetition
Juno tracks covered ground to reduce accidental re-asking. When it returns to a topic, the return is deliberate — approaching something important from another angle.
Quality control
Checked as it happens
Quality checks run while the interview is taking place, not only after it ends — making research discipline part of the conversation itself.
Question-level review
Questions pass through a codified set of qualitative-research checks before they reach the participant.
Bias prevention
Checks look for leading questions, embedded answers, false binaries, fed examples and presuppositions, so problems can be corrected before the question is asked.
Single-question discipline
The interview is designed to split compound questions before they are asked, giving the participant one clear, open question at a time.
Duplicate detection
The conversation so far is considered before the next question, reducing accidental semantic repetition.
Languages
Multilingual interviewing
Research discipline carries across Juno's supported languages, with interviews adapting to how people talk rather than only translating the words.
Text in 100 languages
Juno currently supports text interviews in 100 selectable languages and voice interviews in 9, distinguishing those modalities rather than presenting one ambiguous total.
Dialect awareness
Juno can pick up regional vocabulary from the participant's own answers and adapt its phrasing, so the conversation sounds local rather than generic.
Consistent quality across languages
Common methodology and technique checks apply across supported languages, helping findings remain comparable across markets.
Comparability
Standardised comparison questions
When you need the exact same question asked across every interview, Juno places it word-for-word — without breaking the conversation.
Verbatim questions, placed in context
Approved verbatim questions are woven in word-for-word, at the moment the topic is already warm rather than bolted on.
Placed with judgement
Verbatim questions are prioritised by the time left and topic relevance, and skipped when their premise does not apply to a given participant.
Cross-interview comparability
Identical phrasing supports like-for-like comparison across interviews while the surrounding conversation remains responsive.
Reliability
An interview that holds
An interview is a live conversation with a real person's time on the line. Juno is designed to keep it moving and make degraded runs visible.
Continuity by design
Fallback paths help the conversation recover from service interruptions without asking a participant to start again.
Conversational pacing
Response timing is monitored because an awkward pause does not just test patience — it can change what a participant says next.
Degraded interviews stay visible
When technical problems affect a conversation, the interview is flagged for review rather than quietly treated as normal.
Scoring
Post-interview quality scoring
Post-interview checks score completed conversations, so quality is considered per interview rather than assumed across the batch.
Per-interview quality report
The report records relevance, technique adherence, coherence, research value and participant engagement.
Recorded reasons
The score and its reasons are recorded, giving reviewers evidence for why a conversation passed or needs attention.
Flagged, not folded in
Interviews that fall below quality thresholds are flagged automatically for review rather than slipping into the dataset unnoticed.
Judgement
The mid-interview judgement calls
The judgement calls a researcher makes mid-interview — when to follow, when to let go, when to steer back — Juno makes too.
Participant-led, brief-serving
Juno follows the participant's own language and framing while steering toward the territory the research needs.
Third-party discipline
When others come up — a partner, an employer — they are used for context, then the focus returns to the participant's own decisions.
When a path is blocked
When a participant cannot or will not answer, Juno finds another route or moves on gracefully rather than pressing.
See the standards at work.
Watch Juno run a real interview, read the exchange, and judge the craft for yourself.