Operations, Setup, Meeting Types. Where a meeting becomes a defined object: named, classified by department, given a framework, and told what to look for. Recording happens in the calendar; this is where you decide what a recorded meeting means.
You never write a prompt. You pick a framework, which fills in what to look for in plain language, and you edit that list.
The Two Panes
Left - the meeting-type library, scoped to your department. Admins see it grouped across all departments.
Right - the editor for the selected type, in five sections.
Section 1: Identity
Name the type, classify it by department, and describe in plain language what this meeting is for. The description is not decoration - it goes into the analysis, so "a first call to establish whether there is a real opportunity" produces a different reading than "sales call".
Section 2: Framework
Pick a starting point. The catalogue is per-department: you see your department's frameworks plus the general ones that apply everywhere, and there is always a Custom option that starts blank.
Twenty-five frameworks ship. Choosing one pre-fills the outcomes with the right things to look for, which you can then edit.
| Department | Frameworks |
|---|---|
| Sales | BANT, MEDDPICC, SPIN, Sandler Pain Funnel |
| People | 1-on-1 (GROW), Feedback (SBI), Performance Review, Behavioral Interview (STAR) |
| Engineering | Sprint Standup, Retrospective, Incident Postmortem (5 Whys), Architecture Review (ADR) |
| Marketing | Campaign Kickoff (SMART), Content Review, AARRR, Campaign Review |
| Operations | Process Review (DMAIC), Vendor Review |
| Finance | Budget Review, Board Update |
| Support | Escalation Triage, Customer Feedback (CSAT/NPS) |
| General | All-Hands, General Meeting, 1-on-1 (GROW) |
Tip
Pick the framework that matches how your team already sells or manages, not the most sophisticated one. MEDDPICC on a workspace that runs simple transactional deals produces eight fields nobody fills. BANT is four, and four honest answers beat eight guesses.
Section 3: What To Look For
The outcomes - what the platform extracts from every meeting of this type. The framework fills these in; you edit, reorder, add and remove.
Each outcome has a label, a kind, and guidance telling the analysis how to derive it. The kinds are covered in How Meetings Are Scored.
Write guidance as if briefing a new analyst. "Who controls the budget and can say yes" is guidance. "Economic buyer" is not.
Section 4: Context
What the platform pulls in to coach against real business rather than the transcript alone. Choose any of:
- Deal and pipeline
- Contact
- Company
- Prior meetings
- Open tasks
- Invoices
- Support tickets
You also set a lookback window: 7, 14, 30, 90 or 180 days.
This is the difference between "they mentioned budget concerns" and "they mentioned budget concerns, and this is the third meeting where that has come up while the deal has sat in the same stage".
Keep in mind
More context is not automatically better. Pulling invoices and tickets into a first discovery call adds noise, because there are none. Pull what a good colleague would have read before walking in.
Section 5: After The Meeting
What happens automatically once a meeting is analysed:
| Action | Result |
|---|---|
| Log to deal / contact timeline | The meeting appears in the record's history |
| Create tasks | Action items become real tasks |
| Create engineering issues | Action items become issues |
| Suggest pipeline stage move | A recommendation, not an automatic move |
| Post summary to team chat | The summary goes to the team |
Note that the stage move is a suggestion. The platform does not silently move your deals.
Building A Type: A Worked Example
A discovery call for a sales team running MEDDPICC.
- Identity. Name it "Discovery call". Department: Sales. Describe it: "First substantive call. We are establishing whether there is a real, funded problem and who can buy."
- Framework. MEDDPICC. Eight outcomes appear.
- What to look for. Trim it. On a first call you will not learn the paper process - remove it, or leave it knowing it will usually come back empty. Keep Metrics, Economic Buyer, Decision Criteria, Identified Pain and Champion.
- Context. Deal, Contact, Company. No invoices or tickets - there are none yet. Lookback 30 days.
- After the meeting. Log to the timeline and create tasks. Do not suggest a stage move from a first call.
- Save.
The analysis instruction is composed behind the scenes from the framework, your outcomes and the type's identity.
Common Questions
Can I create my own framework? Choose Custom and define the outcomes yourself. Organisation-specific frameworks can also be seeded for your whole workspace.
Who can edit meeting types? It is a department-scoped library. People see their own department's; senior roles see across departments.
Does this record meetings? No. Recording is set up in the calendar. This defines what a recorded meeting is analysed for.
What if the meeting does not match its type? The analysis follows the type you assigned. A sales call analysed as a retrospective produces nonsense, so assign the type correctly when scheduling.
Did this answer your question?
No, ask a person