Stakeholder reflection — worked example

The stakeholder bundle works best when you use it over time. Here’s what the full cycle looks like in practice, from first setup through a synthesis session before a key meeting.

The scenario

You’re a PM or EM. You’ve just gotten a new manager — let’s call them Trae. You want to build a real picture of this relationship over time, not just a vague sense of “things are fine.”

Step 1: Register

I just got a new manager. Add Trae North to managing-up.

stakeholder-register fires. It walks you through:

Name: Trae North
Category: managing-up
Role and context: Director of Product, your skip-level is the VP of Product. Weekly 1:1s.
Initial notes: New to the role (just promoted), still calibrating his style. Came from a strong IC background.

Confirms before writing: “Here’s the file I’ll create at ~/bettersense-work-reflections/managing-up/trae-north.md. Anything to change?”

You confirm. File is created.

Step 2: Reflect (weekly-ish)

Three weeks later:

Let me reflect on Trae.

stakeholder-reflect loads the file and picks 2-3 questions appropriate to the moment:

“You haven’t answered ‘What does [Trae] seem to be optimizing for that you haven’t explicitly discussed?’ yet. Worth exploring. Things to consider: their stated goals vs. what they’re spending time on, what they escalate vs. let ride, how they respond when your priorities differ from theirs.”

You reflect for 10 minutes. The skill pushes back when you write “he’s hard to read” — “that’s an observation about your experience, not about his behavior. What specifically is happening that makes him feel that way?” — and helps you articulate:

“He asks detailed questions in 1:1s but rarely commits to a direction. In the last three 1:1s, he’s asked me to ‘keep thinking about it’ rather than aligning.”

That’s specific. The skill writes a dated entry with exactly that language.

Step 3: Stay current

Six weeks in, Monday morning:

What's due this week?

stakeholder-due scans all your stakeholder files and surfaces:

Overdue (7+ days):
  → Trae North (managing-up): "What do I appreciate about how Trae works?" — 14 days overdue
  → Sahil Bhatia (managing-across): "What's one assumption I'm making about how Sahil sees my work?" — 9 days overdue

Coming up (due this week):
  → Yulia Krenkova (managing-down): "What's one thing I could do differently in our 1:1s?" — due in 2 days

You open stakeholder-reflect for Trae and answer the appreciation question. Turns out it’s easier to answer than expected — there’s real specificity there you hadn’t articulated. It goes in the file.

Step 4: Synthesize (before a big moment)

It’s been three months. You have a skip-level with Trae’s manager next week, and you want to brief yourself on where things stand.

Before my skip-level next week, synthesize what I know about Trae.

stakeholder-synthesize reads the file — 11 dated entries across 12 weeks — and produces:

Patterns:

Consistently asks clarifying questions before committing — 7 entries reference this. Hypothesis: decision-making style, not indecision. He’s building a map before moving.
Energy visibly higher in conversations about strategy and narrative than execution detail — 4 entries. Worth exploring whether he wants to be consulted on more upstream work.

Contradictions:

Early entries describe him as “hard to read.” Later entries have concrete observations. The gap may be relationship depth, not his style — worth checking if the pattern holds in new conversations.

Blind spots:

No entries in the ‘sense’ category about how Trae is perceived by others. Only your direct experience. Useful to get a second view.

Suggested next:

“What does Trae optimize for that you haven’t explicitly discussed?” hasn’t been fully answered — the entries circle it without landing. One conversation centered on that question could clarify a lot.

That’s the input you needed for the skip-level.

What makes this work

Specificity compounds. One-line entries produce shallow synthesis. Ten specific entries produce a 360-view you couldn’t have assembled from memory.
The skill pushes toward specificity. Every “he seems X” gets challenged: what specifically?
Synthesis is cited. Every pattern in the output references specific dated entries — not a vibe, an artifact.
It composes. After reading the synthesis, you can run:
- /bettersense:one-on-one-prep — for the actual skip-level agenda
- /bettersense:feedback-frameworks — if feedback emerges from the synthesis
- /bettersense:coaching-mode — if a development conversation with a report is what the synthesis points to