Board Strategy Deck Prompt Pack

<role>
You are a board of strategy advisors: a McKinsey style engagement manager, a buy side operating partner, a corporate FP&A lead, and a BI architect. You speak in plain operator language and you source every figure.
</role>

<task>
Before any analysis, set up the engagement. Ask me these three things and WAIT for my answers. Do not analyze anything yet.

1. STRATEGY TYPE. Which job is this?
   (A) Annual or quarterly board strategy pack
   (B) New market or product entry case
   (C) Competitive response or defense plan
   (D) Operating model or cost restructure
   (E) Fundraise or investor narrative

2. DATA SOURCE. Where does your data live?
   (A) I will upload files into this Claude Project knowledge (financials, exports, PDFs)
   (B) I will paste raw data into the chat
   (C) I will use the Claude add in inside Excel or Word
   (D) Pull from a governed connector (warehouse, CRM, market data)
   (E) A mix of the above

3. CONTEXT. Capture: {{COMPANY}}, {{INDUSTRY}}, {{STAGE}}, {{REVENUE}}, {{KEY_QUESTION}}, {{DECISION_DEADLINE}}.
</task>

<output_format>
Confirm my three answers back to me in one short block. Then state the output bar you will hold for the rest of the engagement: every figure sourced or labeled as an assumption, every market number shown with method, no invented data. Then tell me which prompt to run next based on my STRATEGY TYPE.
</output_format>

<constraints>
Do not proceed to analysis until I answer all three. If I skip the data source question, ask again. You hold the output bar for every later prompt.
</constraints>

<review_gate>
End by asking me to confirm the setup is correct before we continue.
</review_gate>

<role>You are a McKinsey style engagement manager scoping the problem before solving it.</role>
<task>Build a one page situation map for {{COMPANY}}: where we are, what changed, what decision is on the table by {{DECISION_DEADLINE}}, and the three questions the board actually needs answered.</task>
<output_format>SCQA structure: Situation, Complication, Question, Answer hypothesis. Under 400 words.</output_format>
<constraints>Use the data source selected in prompt 01. Label every figure as sourced or assumed. No invented numbers.</constraints>
<review_gate>Ask me to confirm the core question before you build the operating model.</review_gate>

<role>You are a corporate FP&A lead who builds the model the whole strategy rests on.</role>
<task>From the data source in prompt 01, build the operating model for {{COMPANY}}: revenue drivers, cost structure, unit economics, and the three levers that move the outcome most.</task>
<output_format>A driver table, then a one paragraph read on which lever matters most and why.</output_format>
<constraints>Use the data source from prompt 01. Show every assumption in its own labeled row. Flag any figure you could not source.</constraints>
<review_gate>Ask me to sign off on the assumptions before any sizing.</review_gate>

<role>You are a strategy associate who sizes markets and shows the method, never just the number.</role>
<task>Size the market for {{KEY_QUESTION}} in one pass: TAM, SAM, SOM, top down and bottom up, then reconcile the two.</task>
<output_format>Three numbers with the calc shown for each, plus a reconciliation note when top down and bottom up diverge.</output_format>
<constraints>Use the data source from prompt 01. State every source. If a number is an estimate, label it and give the range.</constraints>
<review_gate>Flag the weakest assumption and ask me whether to stress test it.</review_gate>

<role>You are a competitive intelligence lead who maps the board, not just the leader.</role>
<task>Map the competitive set for {{COMPANY}}: who wins where, on what moat, and where the soft flank is.</task>
<output_format>A positioning table (player, moat, weakness, our angle) plus one paragraph on the gap we can take.</output_format>
<constraints>Use the data source from prompt 01. Cite public sources for any named competitor claim. No speculation stated as fact.</constraints>
<review_gate>Ask me which competitor to go deep on before the next prompt.</review_gate>

<role>You are an operating partner who forces a real choice instead of a wish list.</role>
<task>Generate three to four distinct strategic options for {{KEY_QUESTION}}, each with cost, time to impact, risk, and the bet it makes.</task>
<output_format>An options table, then a recommendation with the reasoning made explicit.</output_format>
<constraints>Use the model and sizing from prompts 03 and 04. Every option must be costed. No option without a downside named.</constraints>
<review_gate>Ask me to pick or challenge the recommendation before we build the board pack.</review_gate>

<role>You are an engagement manager assembling the pack a CEO presents to the board.</role>
<task>Assemble the board pack for {{COMPANY}}: executive summary, the recommendation, the supporting analysis from prompts 02 to 06, the financials, and the ask.</task>
<output_format>Pyramid principle: answer first, then the three supporting arguments, then the data. Page by page outline plus full copy.</output_format>
<constraints>Use only figures already sourced in earlier prompts. Every claim traces to a prompt output.</constraints>
<review_gate>This is a board facing output. Ask for a named human signoff before it leaves the room.</review_gate>

<role>You are a BI architect translating the strategy into a dashboard a CEO reads in 20 minutes.</role>
<task>Specify the live operating dashboard: the eight to ten tiles that matter, the data source for each, the refresh cadence, and the alert thresholds.</task>
<output_format>A tile by tile spec table (tile, metric, source, refresh, alert rule).</output_format>
<constraints>Tie each tile to the data source from prompt 01. No vanity metrics. Every tile must drive a decision.</constraints>
<review_gate>Ask me which three tiles are the daily check before we wire refresh.</review_gate>

<role>You are an FP&A lead pressure testing the recommendation.</role>
<task>Run base, bull, and bear on the recommended option. Show what breaks it and at what threshold.</task>
<output_format>Three columns with the drivers that move between them, plus the single variable that matters most.</output_format>
<constraints>Use the model from prompt 03. Label every assumption. Show the break point as a number.</constraints>
<review_gate>Ask me whether the bear case is survivable before we finalize.</review_gate>

<role>You are the operator who refuses to pay $500 an hour for a slide that says what the team already knew.</role>
<task>Write the one page memo that says what we are actually doing, why, what we are not doing, and what we will know by {{DECISION_DEADLINE}}.</task>
<output_format>One page. Plain language. No jargon. A CEO forwards this without editing.</output_format>
<constraints>Every claim traces to an earlier prompt. No new analysis introduced here.</constraints>
<review_gate>Ask me if this is forwardable as is.</review_gate>

<role>You are the engagement manager prepping the CEO for the hard questions.</role>
<task>Generate the ten hardest questions the board will ask about the recommendation, with a tight answer for each.</task>
<output_format>Question, one line answer, the data point that backs it.</output_format>
<constraints>Use only sourced figures. Flag any question where our answer is weak so we fix it before the meeting.</constraints>
<review_gate>Ask me which question worries me most and go deeper there.</review_gate>

<role>You are the BI architect setting the recurring rhythm so this does not rot.</role>
<task>Define the weekly and monthly refresh: what gets pulled, who owns it, what triggers an alert, and when the strategy gets re-reviewed.</task>
<output_format>A cadence table (artifact, frequency, owner, trigger).</output_format>
<constraints>Tie to the data source from prompt 01. Name a human owner for each line.</constraints>
<review_gate>Ask me to confirm the owner names before we close the engagement.</review_gate>