The task
Your weekly 1:1 with your manager is in 20 minutes. Last week she opened with “how’s it going?” and you mumbled for 8 minutes before remembering you wanted to bring up the design-review blocker, ran out of time, and rebooked an “offline” that never happened. You have a Notion page of half-finished thoughts from the week — one blocker, one heads-up (an interview at another company), one career question — and you want to walk into the room with three bullets in priority order, your ask in line one, and an answer ready for the follow-up you know is coming.
Where AI helps — and where it does not
AI is excellent at sorting scattered notes into a priority-ranked 3-bullet agenda, predicting the follow-up questions your manager will ask, and proposing both a direct and a softened opener for sensitive topics. It can also flag when you’ve over-stuffed the agenda — six bullets in 20 minutes means nothing gets resolved. What AI cannot do: read the political room. Whether this is the right week to raise comp, whether the recent reorg means you should push back on scope, whether your manager is in a good or bad week — that judgment stays with you.
The named failure mode: the status-recap 1:1. AI defaults to listing what shipped, what’s in progress, what’s next. You walk in, recite status, your manager nods, time runs out, and your real ask never surfaces. Force the prompt: status goes in the async update, the 1:1 is for blockers, heads-ups, and career conversation.
What to feed the AI
- Notes from since the last 1:1 — decisions, friction, wins, questions you parked
- The one thing you want them to know (heads-up: interview, family thing, scope concern)
- The one thing you want from them (decision, approval, advocacy, advice with a specific call)
- Sensitive topics you might raise — comp, promotion, conflict with peer, doubts about role
- Your manager’s known preferences — direct vs warm, agenda-driven vs free-form, morning vs afternoon energy
- The last 2 things your manager flagged for you to follow up on (so AI puts them as opener context)
- Time budget — 15 / 20 / 30 / 60 minutes
- Whether the 1:1 is recurring or skip-level / first-time
Copy-ready prompt
Prep my 1:1 agenda.
Notes since last 1:1 (decisions, friction, wins, parked questions):
{paste}
One thing I want them to know (heads-up): {paste}
One thing I want from them (specific decision, approval, advocacy, advice with a call): {paste}
Sensitive topics I might raise: {comp / promotion / peer conflict / doubts / role change — pick any}
Manager preferences: {direct vs warm, agenda vs free-form, morning vs afternoon}
Last 2 things my manager flagged for follow-up: {paste}
Time budget: {15 / 20 / 30 / 60 minutes}
1:1 type: {recurring / skip-level / first 1:1 with new manager}
Return:
1) Exactly 3 bullet talking points, ordered by priority. Status updates do NOT go here — they go in async. Each bullet must be a blocker, a heads-up, or a career conversation.
2) For each bullet: my one-line ask + my recommendation if applicable + the prep I should bring (data, doc, decision).
3) Two likely follow-up questions my manager will ask + short answers I can prep.
4) For any sensitive topic, give me TWO openers — one direct, one softened — and a one-line cue for when to use each.
5) Opener: 2 sentences max — reference the most recent thing they flagged, then transition into bullet 1.
Rules:
- Status updates do not go in 1:1 — they go in async. Reject any bullet that is "what I worked on."
- Each bullet is a verb the manager can act on ("decide", "advocate", "approve", "advise").
- If I have 6 bullets, cut to 3 and demote the rest to "async after meeting."
- Opener uses what they last flagged — signals "I listened."Shorter variant — single-decision 1:1 prep
Prep a 15-minute 1:1 focused on one decision.
Decision needed: {paste with my recommendation}
Context my manager needs: {3 bullets max}
Likely pushback: {what they might object to}
Output: 2-sentence opener + the decision frame + my answer to the 2 likely pushbacks. Plus a fallback ask if they defer.Sample output
A useful prioritized agenda: “1) Blocker — design review queue is 2 weeks; can we get an exception this week or should I de-scope the v2? (Recommend: exception, the de-scope kills the analytics tier we promised marketing.) 2) Heads-up — I have an offer call Wednesday afternoon at another company; not actively job-searching but want you to hear it from me. Will share details after the call. 3) Career — I want to lead the Q3 platform project. What would I need to demonstrate by EOQ for this to be uncontroversial?”
A useful predicted follow-up: “Manager will likely ask: ‘What’s the impact if we don’t get the design exception?’ Prepared answer: ‘We slip launch by 8 days, marketing has to re-bake the campaign for a partial feature set, and we lose the analytics tier we’d promised. Cost of the exception: design team works one weekend, recoverable.’”
A useful two-opener pair (sensitive): “Direct: ‘I want to talk about whether my comp band reflects the platform-lead scope I’m operating in.’ Softened: ‘I’d like to understand the comp framework for the platform-lead scope I’ve been operating in, especially with the reorg context.’ Use direct if she’s in a steady week and you have evidence. Use softened if you have less data or the company is in a reorg.”
How to refine
- Cut to 3 bullets: “I have 6 bullets. Trim to 3. The other 3 become an async Slack message after the 1:1 — write that message too, 3 bullets max.”
- Reject status creep: “Reject any bullet that is ‘what I worked on’. Status goes in the async update. The 1:1 is for blockers, heads-ups, and career conversation only.”
- Predict the pushback: “For each bullet, predict the one objection my manager will raise and write my 2-sentence answer. If I cannot answer the objection, I’m not ready to raise the bullet.”
- Two-opener for sensitive: “For the comp bullet, give me a direct opener and a softened opener. Add a one-line cue for when to use each. Default to direct if I have evidence.”
- Open with their thread: “Open the meeting by referencing the thing my manager last flagged. Two sentences max. Signals ‘I listened’ before I ask for anything.”
Common mistakes
- Going in with no agenda — the manager drives the full 20 minutes on her topics and your blocker dies unaddressed
- Burying the real ask behind 8 minutes of status — by the time you get to bullet 3, the meeting is over
- Asking for “career advice” with no specific decision attached — managers experience this as work being kicked back to them with no signal of what you want
- Six bullets in 20 minutes — nothing gets resolved, momentum dies for everything
- Same agenda format every week — manager tunes out; vary the opener even if the structure is similar
- Raising sensitive topics (comp, promotion, doubts) without a written ask — the conversation becomes feelings instead of decisions
- Letting the 1:1 become a venting session — feels good in the moment, costs you the meeting’s actual purpose
- Skipping the opener that references their last-flagged thing — signals you don’t listen and undermines your asks
FAQ
- My manager prefers free-form 1:1s.: Send the 3 bullets 30 minutes before as “a few things I had on my mind.” Most free-form-loving managers actually like this once they see how the meeting goes. The structure helps them too; they just don’t want to be the one doing the structuring.
- What if there really is nothing this week?: Skip the 1:1 (with a 2-line Slack: “Light week, no blockers, see you next week”), or use the slot for career conversation. Filler 1:1s erode the value of real ones; manager starts treating every 1:1 as ignorable.
- What if I have a sensitive topic but unsure whether to raise it?: Prep both openers (direct + softened), bring the data, and decide in the room based on her energy. AI helps you arrive prepared; the decision to raise stays human.
- First 1:1 with a new manager?: Different prompt. Open with: 3 things I want her to know about how I work; 1 thing I want to know about how she works; 1 specific recent decision I want her input on. Status comes later.
- What if my manager always runs over and the 1:1 ends with my bullets unaddressed?: Lead with the most important bullet — not in priority of difficulty, but in priority of “if only one thing gets discussed, this.” And ask explicitly at minute 18: “Two more minutes — I want to make sure we get to the platform-project ask.”
Related
- AI manager update emails
- AI promotion self-review
- Behavioral Story Mining Prompts for STAR Stories
- Rewrite a Tense Work Message With AI
- AI Performance Review Draft
- AI Meeting Agenda
Tags: #AI writing #Career #Workflow #Meeting