AI-moderated research in Lookback exists to solve a very specific problem:
how to preserve qualitative depth in unmoderated research when no researcher is present.
It is not about automating conclusions or replacing researchers.
It is about improving data quality during sessions, while keeping humans in control.
The core problem AI moderation addresses
In traditional unmoderated research, common issues include:
participants misunderstanding tasks
participants forgetting to speak aloud
answers that are partial, vague, or incomplete
no way to clarify misunderstandings in real time
These issues do not come from bad participants - they come from the absence of a moderator.
AI moderation exists to close that gap.
AI moderation lives inside Tasks
AI moderation in Lookback operates within unmoderated Tasks.
Researchers provide:
the task or question
the intent behind it
guidance on what a good answer looks like
The AI then uses that context to decide when and how to follow up.
This keeps the researcher in control of:
what matters
what is worth probing
how much structure to apply
What AI does during a session
During an AI-moderated session, the AI may:
ask clarifying follow-up questions
prompt participants to elaborate
encourage participants to speak aloud
ensure questions are fully answered
All of this happens in the moment, while the participant is still engaged.
AI moderation improves the session itself - not just the analysis afterward.
What AI does not do
To be explicit:
AI in Lookback does not:
decide what the findings are
draw conclusions
replace researcher judgment
flatten nuance into summaries
AI assists attention and completeness.
Meaning and interpretation remain human responsibilities.
AI moderation preserves researcher intent
A key design principle in Lookback is that AI follows intent, it does not invent it.
Researchers define:
what they are trying to learn
why a question matters
what signals are important
AI uses this context to behave like a careful assistant - not an autonomous agent.
This makes AI moderation a continuation of the researcher’s thinking, not a substitute for it.
AI-moderated sessions are still full sessions
AI-moderated research produces the same kind of evidence as other methods:
full session recordings
screen, audio, and responses
live streaming as the session happens
findings extracted from real moments
AI moderation changes how the session unfolds, not what kind of evidence is produced.
When to use AI-moderated research
AI-moderated research is especially useful when you want to:
run unmoderated studies at scale
improve the quality of asynchronous responses
reduce follow-up sessions caused by unclear answers
preserve qualitative depth without scheduling live moderation
It works best when paired with:
clear research intent
well-designed tasks
stakeholder goals defined at the Project level
How AI-moderated research fits with other methods
Within a Project, AI-moderated research often complements:
moderated research for deep exploration
unmoderated research for flexibility and reach
Because all sessions stream live and produce comparable evidence, insights remain connected across methods.
Why this matters
AI moderation allows researchers to:
trust unmoderated data more
spend less time fixing avoidable issues
focus attention on interpretation rather than cleanup
It strengthens qualitative practice rather than automating it away.
What to explore next
To continue exploring AI in Lookback:
Learn how stakeholder goals provide context for AI
Explore how Eureka assists analysis during and after sessions
See how AI helps surface value without bypassing evidence