Strategy

How do you turn a valuable problem into choices that can survive reality?

Strategy builds a coherent set of choices that can be tested, changed, and acted on.

It is not a list of goals, projects, or predictions. A strategy explains where to act, how value should be created there, what will not be done, which capabilities make the choice feasible, and what evidence could prove the logic wrong.

Value → Intent → Problem → Question → Strategy → Decision
→ Capability → Control and Action → Proof → Learning → evolved Intent

Choice Record

Use one record for one consequential strategy.

STRATEGY CHOICE RECORD v1

Beneficiary:
Protected value:
Intended outcome:

Observed reality:
Problem:
Strategic crux:

Where to act:
How value will be created:
Exclusions and sacrifices:
Accepted loss until review:

Required human capabilities:
Required agent skills:
Required business capabilities:
Enabling control systems:
Decision authority:

What would have to be true:
1.
2.
3.

Weakest critical assumption:
Smallest test:
Falsifier:
Proof signal:
Review window:

The output is a proposed choice system, not permission to act. Hand a consequential commitment to Navigate a Decision. That method keeps the existing nav.v1 decision contract and records who confirms, revises, or rejects the move.

Build the record

1. Name the beneficiary and intent. State whose agency should increase, which value must be protected, and what observable outcome matters.
2. Face reality and find the crux. Describe the current state, the valuable gap, and the obstacle or opportunity on which success most depends.
3. Create real alternatives. Vary where to act and how value could be created. An option is not strategic if it excludes nothing.
4. Test fit and sacrifice. Name required human, agent, and business capabilities. State the work, market, customer, or benefit you will decline.
5. Ask what would have to be true. Express the logic as assumptions that evidence can strengthen or falsify.
6. Find the weakest critical assumption. Choose the assumption that is both uncertain and capable of collapsing the choice system.
7. Design the smallest decisive test. Freeze the falsifier, proof signal, and review window before acting.
8. Commit through a decision. Give authority and resources to the choice only after its bounds and review trigger are explicit.

Capability and control

Dreamineering integrates three different jobs:

Dreamineering = Intent × Strategy × Capability

• Intent says what future is valuable and for whom.
• Strategy chooses the mutually reinforcing path and accepted loss.
• Capability bounds what is feasible now and what must be built next.

The relationship runs both ways. Current capability constrains credible strategy. Strategy identifies the capabilities worth cultivating. Execution evidence changes the next strategy.

Use the Capability System to keep the owners distinct:

• human capabilities supply questioning, judgment, courage, and strategy-making;
• agent skills provide bounded leverage and a receipt;
• business capabilities make value creation repeatable;
• observed outcomes, not requirements or confidence, promote a capability from required to demonstrated.

Control then turns the committed strategy into governable action: named setpoints, levers, gauges, authority, bounds, and correction.

Venture example

Teaching example — not a field-proven outcome.

A founder believes small building firms lose margin because quoting depends on one experienced estimator.

• Beneficiary and outcome: owner-operators issue reliable quotes faster without surrendering final commercial judgment.
• Crux: can job evidence be structured well enough to prepare a useful first draft?
• Where and how: start with repeat residential renovation work; combine prior jobs, supplier prices, and an estimator review.
• Sacrifice: exclude unusual commercial projects and autonomous price approval.
• Capability fit: human estimator judgment, a bounded evidence-retrieval skill, and a repeatable quoting workflow.
• Weakest assumption: five prior jobs contain enough comparable detail to draft a quote within an acceptable error band.
• Smallest test: blind-draft three historical jobs, then compare with the estimator’s original reasoning.
• Falsifier: the draft misses a declared cost or scope threshold on two of three jobs.
• Review: one week, before customer-facing use.

The strategy is not “build an AI quoting tool.” It is the linked choice of beneficiary, scope, value mechanism, sacrifice, capability, authority, and test.

Journey transfer

This SME Journey is also a teaching example, not a claim about a delivered client result.

An existing wholesaler considering AI-assisted purchasing can reuse the same record:

1. Protect the buyer’s accountability for supplier risk.
2. Diagnose whether late replenishment or excess stock is the valuable gap.
3. Choose one category and one recommendation point.
4. Exclude autonomous purchase approval.
5. Test whether recommendations beat the current reorder rule during a fixed review window.
6. Retain the observed variance before expanding scope.

If the method cannot transfer from venture formation to an existing operation without hiding the different authority and evidence, the record is incomplete.

Evidence boundary

The method combines research-backed ideas with Dreamineering doctrine:

• Roger Martin describes strategy as interrelated choices and uses “what would have to be true” to test strategic options.
• Teece, Pisano, and Shuen connect competitive advantage to organisational processes, asset positions, and paths, supporting the two-way relationship between strategy and capability.
• Gollwitzer’s implementation-intention research supports explicit situation–response plans for translating a chosen intention into action.
• The record, route order, controller question, and teaching examples are Dreamineering doctrine. Their usefulness here remains a proposition until readers apply them and return comparable evidence.

Sources:

• Roger Martin: Strategy
• Roger Martin: The Strategic Choice Structuring Process
• Teece, Pisano, and Shuen: Dynamic Capabilities and Strategic Management
• Gollwitzer: Implementation Intentions

Checks

• Do the choices reinforce one another rather than form a wish list?
• Does every important choice exclude a real alternative?
• Are required and demonstrated capabilities clearly separated?
• Would the falsifier genuinely reopen the strategy?
• Is human authority explicit at commitment and correction points?

Failure modes

• Goals mistaken for strategy: the record says what success looks like but not where or how to act.
• No sacrifice: every audience, channel, benefit, and project remains in scope.
• Capability theatre: required capability is described as already proven.
• Unfalsifiable logic: assumptions can absorb every possible result.
• Action without authority: an experiment starts without a named confirmer, bound, or stop condition.

Context

• Problems — diagnose the valuable gap before designing choices.
• Question Business Strategy — converge on the assumption that could change the choice.
• Positioning — decide the audience, need, difference, evidence, capability fit, and sacrifice.
• Decisions — commit authority and resources.
• Capability System — separate required capability from demonstrated outcome.
• Control — execute within bounds and return evidence.
• Flow — diagnose whether intent, strategy, capability, and attention remain aligned.
• Performance — compare proof with the frozen expectation through the Tight Five.

Questions

Changes my mind: a cold reader cannot complete the record or identify an assumption whose answer would change the strategy.

Next question: Which assumption is uncertain enough to test and important enough to collapse your current choice system?

• What choice would its failure reverse?
• Which capability is required but not yet demonstrated?
• What accepted loss protects the test from scope drift?