Build a business initiative plan

Outcome and preparation

Day 2 · 14:15–15:00 · 45 minutes. Produce a business initiative plan supported by your report. Bring the checked brief, strategy, system overview and templates/business-plan.md. Use 5 minutes for framing, 30 for authoring and 10 to challenge.

A business plan links a customer problem to a feasible operating and financial proposal. An AI-adoption roadmap is only one part. The supplied sources do not establish market size, competitor pricing or customer demand; those must remain research questions.

Step by step

1. Step 1

   State the business decision: a four-week reliability/support pilot, more acquisition spending, or waiting for evidence. Identify the target customer and mark the customer-problem statement as a hypothesis until supported by interviews.

2. Step 2

   Ask Claude to interview you before drafting. Answer using source facts or explicitly labelled assumptions.

   Prompt / exampleCopy

   Help me build a business initiative plan for Northstar from this verified report and strategy.
   Ask the five most important unanswered questions, one at a time.
   Cover customer problem, value proposition, alternative options, delivery capacity and budget.
   Distinguish evidence from assumptions. Do not invent market size, competitors or customer quotes.
   [Attach verified report, strategy and system overview]
   

3. Step 3

   Draft against the template. Require options and trade-offs, operations/work packages, owner roles, dependencies, four-week timeline, assumptions and validation experiments. An owner role is a proposal until a person accepts it.

   Prompt / exampleCopy

   Draft the plan using templates/business-plan.md.
   Compare reliability/support pilot, expanded acquisition and gathering more evidence.
   For each option: expected benefit, cost assumptions, delivery risks, evidence needed and stop criteria.
   Cite supplied facts. Label all proposed budgets, targets and timelines as assumptions pending approval.
   Include a customer interview experiment and the evidence that would change the recommendation.
   

4. Step 4

   Build conservative/base/upside scenarios in a spreadsheet or calculator. Start with the illustrative USD 12000 pilot ceiling, then choose explicit allocations for discovery, engineering, support and review/contingency that sum to it. State that the ceiling is an exercise assumption. Vary cost and benefit assumptions independently; do not present a revenue forecast as established fact.

5. Step 5

   Ask Claude to act as a sceptical CFO. Have it identify unsupported demand, hidden costs, capacity constraints and failure to consider the wait option. Fix the plan, then calculate its totals yourself.

   Prompt / exampleCopy

   Review this plan as a sceptical CFO and CTO.
   Find unsupported facts, missing operating costs, unrealistic capacity and untested customer assumptions.
   Recalculate the scenario totals with a tool. Identify one experiment that would invalidate the preferred option.
   Do not approve expenditure. Return specific corrections and remaining decision questions.
   [Paste draft plan and scenario table]
   

6. Step 6

   Save handover/business-plan.md and the scenario table. End with a decision requested, named reviewer to confirm and what evidence is needed next—not an invented board approval.

Expected output and verification

A plan covering customer, problem, value proposition, options, operating model, timeline, budget, scenarios, risks and validation. It should explain why an operational signal matters for a business decision without asserting that the signal proves a causal business impact.

Verify each budget sum, distinguish cost from benefit and note any source date. Example allocation of the ceiling: discovery 2000 + engineering 6000 + support 2000 + review/contingency 2000 = 12000. This is illustrative, not a vendor quote or financial commitment.

Troubleshooting

If the output is just an AI rollout list, ask for the customer, commercial choice and operating model. If market figures appear without a source, remove them and write a research task. If every scenario is profitable, challenge the assumptions and include a case where benefits fail to appear.

Completion checklist

• Business plan and three scenarios saved.
• Budget arithmetic checked; facts and assumptions labelled.
• Customer validation experiment and decision gate included.