From sales calls to product insights, one checked step at a time.

1. Record what the call established

Start with notes you can attribute to the conversation. Record the account, participant role, date, situation, stated problem and any commitment. Mark paraphrases and your own interpretation. Keep private data within the source permissions you have.

Sales needs to know what happens next in the opportunity. Research needs to know which experience supports a claim. Product needs a decision question. One broad summary is unlikely to serve all three equally well.

Copyable working template

Source: call-note-01 (synthetic)
Account: A | Role: operations lead
Observed situation: checks reports before sending
Stated problem: does not know who approved the last report
Commitment: none
Unknown: approval owner and frequency of the problem

2. Resolve the deal's immediate next step

In Founder Sales, review the interaction extraction and proposed opportunity changes. Correct any inferred commitments before accepting a state change. Draft a follow-up from accepted facts and inspect the message before sending it.

The right next step in this example is a clarifying question about approval, not a promise to build a feature. Record an actual send separately; a saved draft or a completed workflow does not prove delivery.

3. Turn the request into an attributed research observation

Select the permitted source for Customer Discovery and choose a research objective. Review the extraction against the notes. Keep the requested solution alongside the situation and consequence that prompted it.

Do not flatten a sales objection into a universal customer need. A buyer asking for approval software and a user struggling to verify a report may be describing different problems.

4. Check the apparent pattern across independent sources

Select the interviews that match the segment and time scope. List distinct accounts as well as conversations. Inspect contradictions and evidence that weakens the proposed pattern.

Synthetic continuation: account B also checks reports but has no approval owner. The common wording does not establish a common workflow. Keep both interpretations alive until another conversation or observation distinguishes them.

5. Ask a bounded product question

Bring the selected evidence into Product Decisions with an explicit question: should the next experiment test report verification or approval ownership? Compare a manual checklist and a formal approval step against the evidence and your constraints.

Inspect the critique and make the required founder decision. If the evidence trace is missing, return to the sources. You can defer a choice and request another interview instead of accepting a recommendation prematurely.

Copyable working template

Decision question:
Options and constraints:
Evidence supporting each option:
Contradictions and missing evidence:
Chosen experiment and owner:
Observation that would change the decision:

6. Keep the handoff explicit

For a manual handoff, choose the source or artifact, retain its reference and tell the next team which question it should answer. Supported cross-Room routes retain their own scope and human decisions; this guide does not assume a universal automatic chain.

The example is synthetic and contains no measured customer result. In a real task, judge each result against its input, record whether it was useful and check the next observed action. Use that record to decide whether this workflow deserves another run.

Continue with a related workflow

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.