Unmoderated research in Lookback is still qualitative research.
It is not about clicks or completion rates - it is about understanding why people do what they do, even when no researcher is present.
This article explains how unmoderated and AI-moderated research works in Lookback, when to use each approach, and what makes it different from typical unmoderated testing tools.
WHEN TO READ THIS
Read this article if you are:
• planning task-based or self-guided research
• concerned about data quality in unmoderated studies
• comparing unmoderated research tools
• exploring AI-assisted follow-ups
UNMODERATED DOES NOT MEAN UNQUALITATIVE
A common misconception is that unmoderated research only captures outcomes.
In Lookback, unmoderated sessions record:
• screen
• audio
• camera (except when researching native app on iOS)
Participants are encouraged to speak out loud, explain decisions, and reflect as they go. The result is rich qualitative material - not just task success or failure.
TASKS VS SELFTEST
Lookback supports two unmoderated modes.
Tasks
Tasks are used when you want structure and comparability.
Tasks allow you to:
• create step-by-step flows
• mix task types (spoken, text, single-choice, multiple-choice)
• add follow-up questions
• randomize tasks
• create unrandomized blocks within randomized sequences
Tasks are ideal when you want to scale qualitative research while preserving intent.
SelfTest
SelfTest is a lighter-weight option.
• minimal structure
• fewer prompts
• faster to set up
SelfTest works well when you want open exploration or early signal rather than comparability.
WHY AI MODERATION EXISTS
The biggest weakness of traditional unmoderated research is simple: No one is there to course-correct.
Participants may:
• misunderstand a question
• forget to think out loud
• only answer part of what was asked
AI moderation is designed to address this specific problem.
AI-MODERATED FOLLOW-UPS
In Lookback, AI moderation lives inside Tasks.
Researchers:
• enable AI follow-ups
• provide context and intent for each question
During the session, the AI:
• asks clarifying follow-ups
• prompts participants to elaborate
• helps ensure questions are fully addressed
The researcher defines intent.
The AI helps preserve it at scale.
This improves data quality without removing the human from the loop.
LIVE STREAMING OF UNMODERATED SESSIONS
A unique aspect of Lookback is that unmoderated and AI-moderated sessions stream live to the dashboard as soon as they start.
This allows teams to:
• observe early sessions
• catch issues quickly
• start sense-making before data collection ends
This behavior is often unexpected - and highly valuable.
DEVICE-SPECIFIC CONSIDERATIONS
Some behavior varies by device.
For example:
• on iOS, when participants go to a native app Lookback can no longer record their camera - you will only get screen and audio
• camera recording pauses if the participant backgrounds the app
These constraints come from operating system policies, not Lookback design.
Previewing on target devices is always recommended.
COMMON MISTAKES TO AVOID
• assuming unmoderated research cannot capture “why”
• skipping previews before launch
• over-structuring when exploration is needed
• under-structuring when comparability matters
The goal is not automation - it is usable qualitative evidence at scale.
WHAT TO READ NEXT
• Working With Stakeholders in Qualitative Research – for impact and collaboration
• Working With Recordings, Findings, and AI – for analysis and synthesis
• Task Setup and Configuration – for step-by-step instructions