Keep a record for each observation
Start with the source rather than a theme label. Record an observation, its participant or account reference, segment, date and situation. Separate a quoted statement, a paraphrase and your interpretation.
Use a permitted internal reference for private material. Avoid copying personal details into every summary. The record should let an authorized reviewer return to the source without spreading the source across unrelated workspaces.
Copyable working template
Observation ID:
Source reference and date:
Account reference and segment:
Situation / task:
Observed problem and consequence:
Requested solution:
Interpretation (separate):
Permission / intended research scope:Compare like with like across weeks
Choose the segment and period before cross-interview synthesis in Customer Discovery. Keep repeated conversations with one account linked, so a vocal account does not become many independent votes.
Ask whether the underlying workflow changed since the previous conversation. A new source can update your understanding without erasing the older record and its context.
Preserve the exception beside the pattern
A candidate pattern needs its supporting accounts and a contrary example. Label unknown scope rather than assigning a confidence percentage that the data cannot support.
Synthetic example: three notes mention slow reporting. Two describe manual data checking; one describes waiting for a manager. Keep those causes separate instead of naming all three an approval problem.
- Candidate pattern
- Reporting requires additional work before sharing.
- Possible cause A
- Manual verification of source figures.
- Possible cause B
- Waiting for another person's approval.
- Evidence gap
- No observed workflow yet establishes which delay matters most.
Review one research gap at a time
A weekly review can end with one next interview rather than a longer feature backlog. The useful output is what you will test and why.
- Inspect source attribution and account counts.
- Check whether a requested solution replaced the original problem.
- Keep changed segment or time scope explicit.
- Choose a question that distinguishes the competing explanations.
- Select the evidence to pass to Product Decisions only when a decision question exists.
Keep working summaries separate from accepted truth
AI themes and room summaries help you find and organize work. They do not independently author your company's canonical facts. A founder reviews a material interpretation before using it to guide product work.
The record template and reporting example are synthetic and carry no claim about real customer counts. Use your own authorized sources, retain their provenance and check that a later summary still represents them accurately.
Access and review
Closed Beta requires an application, manual approval, invitation and verified login. Applying does not grant immediate access. Six ready-made teams, guarded Custom teams and the Apps catalog are included. No Oqera payment or subscription is required. Bring your own AI provider key and pay the provider directly. External actions require specific permission and review; unattended schedules and triggers are off.