E2 — Team One-to-Ones — Meeting Prep
Sessions: week of 2026-07-08 · Prep started: 2026-07-06 Team: 2–5 people, roles/accounts not yet known to Amir. Reframed per Amir's own instinct: optimise for the people map, not information extraction.
PrAxis workshop. Exploratory by design — not the Executive Brief yet.
Stage 1 — Understand the Meeting(s)
True purpose — two layers, not one. The surface layer is the functional download the Weekly Loop already names: accounts owned, capabilities, how work actually happens, whitespace. The deeper layer — and the one worth optimising for, per your own read — is the people map: who creates energy, who creates clarity, who people naturally follow, who is under-utilised, who understands customers versus process, where institutional memory really sits, who is quietly influential despite a smaller title. The functional questions aren't separate from this — they're the vehicle. You're watching how each person answers as much as what they answer.
What success looks like. After five conversations: the account book substantially on paper (surface layer), and a working, provisional answer to each of the seven people-map questions above (deep layer) — even if some answers are still low-confidence after round one.
What can only come from each individual, one-to-one. Their own self-narrative — strengths, what they want from this. Their honest read on team dynamics, which they will not offer in the group setting. Their private view of who on the team really understands customers versus process — a mildly political read on peers that only surfaces in private. This is exactly the kind of signal the joint team meeting structurally cannot produce.
What should not happen in these conversations. - Don't ask "who's the most influential" or "who do you follow" as a direct question. Too blunt this early — it reads as a loyalty test, not a genuine question, and the honest answer needs to be inferred across five conversations, not extracted from any one. - Don't evaluate or rank anyone to their face — the First-Week Plan is explicit that week one is not an audit. - Don't relay what one person said about another. The map depends on cross-referencing five private accounts; the trust that produces honest answers depends on none of that content crossing back.
What a first-class executive does differently. An average manager collects facts and closes the loop — accounts, capabilities, done. A first-class executive listens for energy shifts (which accounts or topics animate someone versus which flatten their voice), who they credit when explaining how something works, whether they narrate problems as "we" or "I" or point elsewhere, and whether they speak with more fluency about customers or about internal mechanics — itself diagnostic of orientation. They also read absence: who doesn't get mentioned unprompted, which accounts nobody claims, and treat both as signal, not just gaps to fill later.
Fact / assumption / hypothesis / political-signal split
| Type | Item |
|---|---|
| FACT | 2–5 direct reports; roles and account ownership not yet known to Amir. |
| ASSUMPTION (untested) | That whatever hierarchy shows up in the joint team meeting predicts individual dynamics. Likely wrong in places — someone quiet in the room may be highly capable or quietly influential; that's precisely the CLAUDE.md-flagged risk of reading only the group setting. |
| HYPOTHESIS | Whoever understands customers best on this team is not necessarily the most senior or most vocal person. Worth actively testing across the five conversations rather than assuming seniority correlates with customer fluency. |
| POLITICAL SIGNAL to watch for | Whether people mention colleagues spontaneously or only when prompted, and whether anyone consistently undersells or oversells their own or a peer's role — early read on internal competitive dynamics or protectiveness. |
Decided: one shared architecture, per-person section within it.
Stage 2 — Conversation Architecture (shared spine, run once per person)
Themes
| Theme | What it's really about |
|---|---|
| A. Who they are | Background, strengths, how they like to work, what they want from this. |
| B. Their accounts | Ownership, history, pain points, commitments already made. |
| C. How work actually gets done | Where it breaks, informally — not the org chart's version. |
| D. Whitespace | Under-served or never-pursued accounts that would fit the differentiated portfolio. |
| E. Silent, not spoken | The people-map read — energy, who they credit, customer-fluency vs. process-fluency, institutional memory, spontaneous (vs. prompted) mentions of colleagues. |
Flow (same order each time, for comparability)
Personal, low-stakes opening (A) → their accounts (B) → how work really happens (C) → whitespace and tacit knowledge (D) → close with two open, low-pressure questions: "anything else useful for me to know" and "what do you want from this working relationship."
The cross-person signal tracker
Update this after each conversation, not just at the end — later conversations should actively test hypotheses the earlier ones raised, not just add five parallel static reads.
| Person | Creates energy | Creates clarity | Naturally followed | Under-utilised | Customer-fluent vs. process-fluent | Institutional memory | Quietly influential |
|---|---|---|---|---|---|---|---|
| (fill after each session) |
Branches
- Guarded, short answers: don't force it — note the guardedness itself as a data point (low trust, or just personality; needs a second read later), don't over-interpret from one conversation.
- Very forthcoming, volunteers views on colleagues unprompted: valuable, but hold what they say about others as
[HYP], not[FACT], until a second person corroborates it independently — don't let the most talkative person frame your view of the rest of the team before you've met them. - Overlapping or conflicting account ownership between two people: don't resolve it on the spot in either 1:1 — resolving it based on one person's account risks looking like you've taken a side. Log as open and cross-check later.
- A grievance about a colleague surfaces: listen without validating or piling on; log it, then actively seek independent corroboration rather than letting it colour the map on one source.
Difficult questions (softened, not evaluative)
- Rather than "who's most influential" — ask topic-anchored versions: "As I build out the operating model, whose input on [X] will matter most?" Same signal, framed forward-looking rather than as ranking peers.
- Reserved for later conversations in the round, once some trust is established: "Is there anything currently going unsaid on this team that I should know?" — the highest-value, highest-trust-required question; don't lead with it in the first 1:1 of the round.
Political considerations
- Order effects. Whoever you talk to first anchors your initial mental model of the team. Be deliberate about sequencing, not just diary-convenience — see the open question below.
- Cross-referencing discipline. Where two people's accounts conflict, this genuinely belongs in
Political_Economy_and_Decision_Rights.mdQ10 (informal influence) as an open item, not resolved by majority vote of two. - Confidentiality in a small team. With only 2–5 people, a whitespace finding or an account-ownership note is often attributable to a single source even without a name attached. Handle downstream use (Discovery Pack, Shashi conversations) with that in mind.
Stage 3 — Pressure Test
- Missing: no built-in checkpoint to synthesise mid-round. Added above (update the tracker after each session, not at the end) — otherwise five readings just sit in parallel instead of testing each other.
- Sequencing risk, real and unresolved: if the most senior or most vocal person goes first, their framing risks anchoring how you read everyone who follows. Worth deliberately not starting with whoever is most senior/dominant, unless there's a scheduling reason that overrides it. Open question below — need your input on relative tenure/seniority to sequence this properly.
- Assumption to challenge: that the five team members have roughly equal standing. Untested — likely a mix of tenure and seniority that should shape both sequencing and how much weight to put on any one person's read of "how work really happens."
- Premature: don't try to resolve account-ownership conflicts or propose any team/org changes in these conversations — week one is listening, not restructuring.
- Risk: over-fitting the people map after one round. Five conversations produce
[HYP]s, not[FACT]s — hold them at appropriately low confidence, consistent with the workbook's own claim-tag discipline, until a second round or real events confirm them. - Missed opportunity: the closing question ("what do you want from this working relationship") is a genuine forward-looking commitment device — worth logging per person as an input to how you design the ongoing 1:1 cadence, not just filed as a nice answer.
Sequencing: not yet known — flagged as a live decision to make once tenure/seniority is clearer, ideally before the round starts. Default rule of thumb if it's still unclear on the day: don't let diary convenience alone decide who goes first; if a natural senior/vocal figure emerges, hold them for the back half of the round.
Stage 4 — Executive Brief
One brief, reused across all five conversations — reread before each session, not just once.
Meeting Objective Get the account book onto paper, and build a working, low-confidence people map across the round — who creates energy, who creates clarity, who's naturally followed, who's under-utilised, who's customer-fluent vs. process-fluent, where institutional memory sits, who's quietly influential.
Critical Questions 1. Who are you — background, strengths, how you like to work, what you want from this? 2. Which accounts do you own — history, pain points, commitments already made? 3. How does work actually get done, and where does it break? 4. Who have we walked away from, under-served, or never pursued that would fit the differentiated portfolio? 5. What do you want from this working relationship going forward?
Things to Listen For - Energy shifts — which accounts or topics animate them versus flatten their voice. - Who they credit, unprompted, when explaining how something works. - Customer-fluency versus process-fluency — do they light up describing the customer, or the mechanics? - Who they mention spontaneously versus only when asked directly.
Political Signals - Whether their account of "how work gets done" matches or diverges from what colleagues will say — don't resolve conflicts live, log them. - Any grievance about a colleague — listen, log, corroborate independently before it shapes the map. - Whether they undersell or oversell their own role relative to what you'll hear from others.
Watch Outs - Don't ask "who's most influential" directly — use topic-anchored versions instead. - Don't audit, rank, or evaluate anyone to their face. - Don't relay what one person said about another. - Don't resolve account-ownership conflicts on the spot. - Update the signal tracker after each session, not just at the end.
Definition of Success By the fifth conversation: the account book substantially on paper, and a first-pass, honestly-low-confidence answer to each of the seven people-map questions — ready to be tested, not treated as settled, over the following weeks.
Not done yet, by choice: Stage 5 (debrief framework) — parked, consistent with E1 and the team intro meeting.