Lookback is designed around a simple idea: qualitative research is about understanding why, not just collecting outputs.
To support that, Lookback organizes research in a way that reflects how real qualitative learning happens over time - not as isolated tests, but as an ongoing process of inquiry, evidence, and shared understanding.
This structure is intentional. It shapes how research is conducted, how evidence is interpreted, and how insights remain useful long after a single session ends.
Research is organized into Projects, Rounds, and Sessions
At the highest level, Lookback organizes research into three structural layers:
Projects
Rounds
Sessions
Each layer serves a distinct purpose. Together, they form a model that supports iterative, qualitative learning rather than one-off studies.
This structure exists independently of specific methods, devices, or workflows. It is the backbone of how research is organized and how evidence accumulates over time.
Projects represent what you are trying to understand
A Project is the long-lived container for a research question, product area, or initiative.
Projects are designed to persist. They are not tied to a single study or test. Instead, a Project represents the thing you are trying to understand β for example, a product, a feature area, or a strategic problem.
All evidence ultimately lives at the Project level. This allows insights to accumulate, connect, and remain relevant as research evolves.
Rounds represent how you are learning right now
Within a Project, research is conducted in Rounds.
Rounds exist to support iteration. They allow you to:
use different research methods
test different variants or prototypes
work with different participant segments
revisit a question as understanding evolves
Importantly, Rounds do not separate findings. This is by design.
Rounds describe how you are learning at a given moment - not what you ultimately learn. Keeping findings at the Project level ensures that insights remain connected, comparable, and cumulative across time.
Sessions represent individual interactions with participants
A Session is one completed interaction with a participant.
Sessions are the atomic unit of qualitative evidence in Lookback. Each session captures what happened during that interaction - such as video, audio, screen recordings, notes, and timestamps - depending on the research method and device.
Everything else in Lookback builds on Sessions. Findings, themes, reels, and AI-assisted analysis all originate from session data.
Research involves different people - with intentionally different experiences
Qualitative research is inherently multi-perspective. Lookback is therefore designed around three distinct human roles, each with a purpose-built experience:
Participants, who provide authentic input
Researchers, who guide inquiry and make sense of evidence
Stakeholders, who observe, align, and build shared understanding
These experiences are intentionally asymmetric.
Participants do not see observers.
β
Stakeholders cannot disturb or directly influence sessions in real time.
β
Researchers work from a control room designed to reduce cognitive load.
These differences are not limitations. They exist to protect research quality, ethics, and evidence integrity.
Why this structure exists
This model reflects how qualitative research actually works:
Learning happens over time, not in single tests
Evidence gains meaning through comparison and context
Insights are strongest when they remain close to raw material
Research has greater impact when stakeholders can follow the trail from question to evidence
By separating structure (Projects), iteration (Rounds), evidence (Sessions), and perspective (roles), Lookback makes it easier to reason about research without losing nuance or continuity.
Not a meeting tool. Not a survey tool.
Lookback is not organized like a video conferencing tool, where sessions are isolated meetings.
It is also not organized like a survey platform, where responses are reduced to rows and columns.
Lookback is a qualitative research system - one where video-based evidence, human judgment, and iterative sense-making remain central, even as AI helps reduce cognitive load.
What to explore next
To go deeper into how Lookback works:
Learn what a Project represents in practice
Understand how Rounds support iteration without fragmenting insights
Explore what a Session contains and what is recorded
See how different research roles experience Lookback by design
Each of these concepts is covered in more detail in the following articles.
