These are the prompts we’ve watched COOs, VPs of Operations, and GMs actually use to run their businesses better — drawn from real client patterns at founder-led, family-run, and PE-backed companies.
Each one is a starting point, not a finished product. They work best when you:
- Modify them for your specific business context
- Add your own data sources to enrich the responses
- Run them inside a Claude Project (or equivalent) with your business knowledge loaded
Copy a prompt, paste in your data where the brackets are, and iterate. Don’t expect perfection on the first run.
Weekly business reviews
What it’s for: Turn a week’s worth of scattered operational data into a one-page leadership briefing you can read in five minutes.
Synthesize the following weekly operational data into a 1-page leadership briefing. Structure:
- Top 3 wins this week (specific, with numbers)
- Top 3 issues requiring attention (specific, with proposed next actions)
- Operating metrics movement (vs. last week and vs. trailing 4-week average)
- 1 observation that wouldn't show up in the metrics alone
Data to synthesize:
[paste weekly metrics, deal updates, customer notes, team updates here]
Style: direct, written for a senior leader who reads it in 5 minutes. No filler. Lead with the most important.
What it’s for: Build the operations section of a quarterly business review before you open a slide deck.
I'm preparing for a quarterly business review with the leadership team. Help me draft the structure for the operations section.
Context for this quarter:
[paste high-level quarter context — revenue, key initiatives, headcount changes, customer notes]
Operational data:
[paste operational metrics]
Draft a 6-slide outline covering: quarter summary, KPI movement vs. plan, operational wins, operational concerns, next-quarter priorities, ask of the leadership team. For each slide, list the 3-5 specific points to make and the data to show.
Customer and account analysis
What it’s for: Get a conservative, evidence-backed read on whether a specific account is at risk.
Analyze the following customer's recent activity and flag any concerning patterns. Pay attention to:
- Order frequency and volume trends
- Communication tone and frequency
- Payment behavior
- Service tickets or escalations
- Any explicit feedback (positive or negative)
Customer data:
[paste customer activity for last 90 days]
Output: 1-paragraph health assessment, 3 specific data points supporting the assessment, 2 specific recommended actions for the account team. Conservative read — don't over-alarm if the data is ambiguous.
What it’s for: Score and triage upcoming renewals so the account team knows where to spend its attention.
For each customer in the attached list with a renewal in the next 90 days, score the renewal risk on a 1-5 scale (5 = very likely to renew, 1 = very likely to churn). Cite 2-3 specific data points supporting each score. Flag any score of 3 or below for account-team attention with a specific suggested intervention.
Customer list:
[paste customer list with usage data, payment history, recent interactions]
Workflow analysis and improvement
What it’s for: Pinpoint where a workflow is actually stuck — and what data would confirm it.
I'm trying to identify the bottleneck in our [workflow name]. Here's how it runs today:
[paste workflow steps, time per step, owner per step, handoffs]
And here's where the team has flagged pain:
[paste team observations]
Analyze where the bottleneck most likely is, what evidence supports that diagnosis, what alternative diagnoses to consider, and what specific data would confirm or refute the primary hypothesis. Be specific. Don't hedge.
What it’s for: Honestly assess whether a given workflow is worth automating with AI — and roughly size the value and cost.
For the workflow described below, assess whether it's a good candidate for AI automation. If yes, draft a 1-paragraph scope including:
- What AI would do in the workflow
- What humans would still do
- The data sources needed
- The integration level required
- A best estimate of annual run-rate value
- A best estimate of build cost
- Top 3 risks
Workflow:
[paste workflow description, volume, current pain, current people involved]
Be honest if AI isn't appropriate here — recommend a simpler tool or a manual fix if that fits better, with reasoning.
Decisions and tradeoffs
What it’s for: Sharpen a hard decision into a frame you can act on — without the model deciding for you.
I'm weighing the following decision: [decision statement].
Options I'm considering:
[list options]
Help me build a structured decision frame. For each option:
- What's the expected value if it works (best estimate)?
- What's the downside if it doesn't?
- What's the probability of success (your best guess)?
- What's the reversibility (easy to undo, hard to undo, irreversible)?
- What information would change my mind?
Don't pick the answer for me. Sharpen the frame so I can decide.
What it’s for: Map how each stakeholder group will react to a change before you announce it.
We're rolling out [change]. I want to map stakeholder reactions before we move.
For the following stakeholder groups, predict the likely reaction (supportive, neutral, resistant, blocking), the underlying concern driving the reaction, and the specific message that would land best with each:
Stakeholders:
[list — e.g., field operations, customer service, sales team, IT, key customers]
Context on the change:
[paste change description, timeline, who's affected]
People and team
What it’s for: Organize your thinking before a candid performance conversation.
I'm preparing for a performance conversation with [person/role]. Help me organize my thinking.
What I know:
[paste recent performance observations, both positive and concerning]
What I want to communicate:
[paste your intent for the conversation]
Help me structure: opening, 2-3 specific positive observations with examples, 2-3 specific developmental observations with examples, the ask (what I want them to do differently), and how I'll support them. Tone: direct, candid, supportive. Not corporate.
What it’s for: Draft a tight weekly team-meeting agenda that doesn’t pad to fill the hour.
Draft a 60-minute agenda for our weekly operations team meeting. The team is:
[list roles]
This week's context:
[paste 2-3 sentences on what's happening this week]
The agenda should include:
- 5-min opening (specific opening question)
- 30-min main content (the 1-2 things most worth discussing)
- 15-min individual updates (3-min each, with a specific prompt for each role)
- 10-min closing (specific closing question and ask)
Don't pad. If we don't have 60 minutes of content, build a 45-minute agenda.
Vendor and process management
What it’s for: Produce a one-page, no-spin evaluation of a vendor you’re considering.
We're evaluating [vendor] for [purpose]. Based on the conversation notes and materials below, draft a 1-page evaluation summary:
- 3 strengths (with evidence)
- 3 concerns (with evidence)
- 3 open questions to ask in the next call
- A go/no-go recommendation with reasoning
Notes and materials:
[paste vendor materials, call notes, references]
Don't soften concerns. Don't pad strengths.
What it’s for: Turn tribal knowledge into a written process a new hire can actually follow.
Draft a documented process for [workflow]. Audience: a new hire learning this workflow in their first 30 days.
Sources to use:
[paste any existing notes, transcripts of how people describe it today, observed steps]
Structure:
- Purpose (1 paragraph)
- Inputs and outputs
- Step-by-step (numbered)
- Decision points and routing rules
- Common edge cases and how to handle them
- Where to escalate
- Quality check at the end
Style: written for someone learning, not someone auditing. Plain language. Specific examples.
A note on customization
These prompts are starting points. The best version of each is the one tailored to your specific business, with your knowledge and your style loaded. Don’t use them as-is for the long term — modify, refine, and build your own library over time.
If you’d like help building a serious internal prompt library and the workflows around it, that’s the kind of work an AI Office retainer covers in its first few months. The 30-minute intro call is the right place to start.