A useful plan is not a record of everything imaginable. It records consequential decisions, uncertainty and the next action in a form that someone can execute.

From business model to execution plan

A Business Model Canvas explains how we expect to create and capture value; an execution plan addresses what must be ready, who will do the work, with which resources, in what sequence and with which acceptance evidence. Nor is it identical to a business plan for investors: the latter may describe market, team, strategy, funding and financial projections in greater detail. The SBA distinguishes traditional and leaner plans; neither is mandatory for every organization [R07].

Use the chain evidence → decision → deliverable → activity → owner → date → control. For instance, if preliminary interviews indicate that offices need lunch before 1 p.m., do not record merely “be punctual.” Document the provisional delivery-window decision, the deliverable “delivery process tested,” the activities needed, the tester and the punctuality criterion. A preliminary observation remains contextual evidence, not a guaranteed commercial result.

Which documents does a business actually need?

A solo business may begin with one shared folder, a governing note and several linked spreadsheets. A regulated operation with several teams may require more formal controls. The goal is not to maximize the document count, but to give each important question an unmistakable home.

Question Document Minimum content
What outcome are we preparing? Scope and milestone sheet Outcome, boundaries, acceptance
What must be delivered? WBS and dictionary Code, work package, exclusions
Who decides and delivers? RACI Roles and ownership
In what order and when? Schedule Duration, predecessors, calendar
With what funding? Budget and cash flow Amount, payment/receipt date, source
What must we comply with? Requirements register Authority, rule, evidence, renewal
What could go wrong? Risk register Cause, consequence, response, owner
Are we ready? Launch readiness checklist State, evidence, approver

Avoid copying all the details of one register into the others. Use stable identifiers: DEL-03 for a deliverable, ACT-08 for an activity, REQ-02 for a requirement and RISK-04 for a risk. Then a schedule can refer to ACT-08 depending on REQ-02 without reproducing the underlying regulation.

To distinguish who performs a deliverable, who accepts its result, and who needs to participate in the decision, use the RACI matrix. A document index listing owners does not by itself resolve ambiguous approval authority.

Step by step: gather, convert and record decisions

First, recover earlier learning. Take the problem and participants from opportunity assessment, price and cost restrictions from business model scenarios, and actual findings from the experiment log. Label each statement as observed evidence, pending assumption or provisional decision.

Second, define a limited initial delivery. “Launch a catering company” is vague; “be capable of a controlled pilot of up to 30 lunches a day in one area under verified operating conditions” is a planable outcome. Use milestones and priorities to specify what is included and excluded.

Third, break down before assigning dates. A work breakdown structure keeps purchasing, training, permissions, safety and support from being hidden behind a single “launch service” task. Assign owners and establish the logical sequence afterwards. Do not start with “two months” and force estimates to match the desired date.

Fourth, establish a review rhythm. In a short weekly meeting, or a solo check-in, ask: what was finished with evidence, what blocks the next step, which assumptions changed, which decisions need recording? Scope, cost or schedule changes should capture who decided, when and why; minor day-to-day adjustments do not require more bureaucracy.

Example: Almuerzo Cerca's preparation file

The governing note reads: objective: a controlled pilot of up to 30 daily orders; location and operating requirements not yet established; target date to be assessed after permitting and purchasing lead times are known. Its linked registers include an operations WBS, preparation budget, requirements research and risk map. Potential authorization is labeled “to verify,” never “obtained,” and no particular permit is invented without a jurisdiction.

Decision Current basis State Next action
Offer a delivery window Hypothetical preliminary research Provisional Run delivery rehearsal
Prepare 30 units per day Capacity assumption Unverified Measure production and delivery
Use an external kitchen Operating alternative Open Obtain quotes and investigate requirements

Mistakes and quality checks

Avoid a purported single source of truth that is out of date, numbers without provenance, decisions with no owner, and links to spreadsheets nobody maintains. Mark estimates as estimates, retain versions, and do not mistake a detailed schedule for certainty about the future. A good project file allows a newcomer to discover what is intended, what is known, what is missing and the next decision.

Use a master execution plan and decision log

The master plan is an operating index, not a duplicate of every file. It identifies the current version of each record, its owner, and the decision it supports. For Almuerzo Cerca, maintain six connected records: researched opportunity, business model, hypotheses and experiments, WBS and responsibilities, resource-loaded schedule, and budget with requirements and risks. A small project may manage these as tabs in one workbook; a complex project may require separate linked documents.

Traceability matters: a channel appearing in the Business Model Canvas does not mean a supplier has agreed to deliver it, and interest recorded in an experiment does not demonstrate paid demand. Keep observed evidence, assumptions, decisions, and pending actions in separate fields. Link each decision to the original evidence, including its limitations.

ID Source Provisional decision Evidence status Owner / next action
DEC-01 Pillar 1: office-lunch opportunity Test one small service area Exploratory signals, not proven demand Owner: define the area
DEC-02 Pillar 2: ordering model Use manual order capture for the pilot Scope decision Owner: test the record
DEC-03 Pillar 3: experimental results Do not assume willingness to pay Paid-demand test missing Owner: plan subsequent evidence
DEC-04 Financial plan Separate PEN 12,500 in payments from PEN 3,000 in cash Illustrative assumption Owner: request quotations

A consequential change needs an actual version, date, author, rationale, affected deliverables, and approval decision. Changing the delivery area may require updating the scope, WBS dictionary, routing, budget, risks, and readiness checks. It must not create six contradictory copies of one assumption. Give pending questions an owner and review date; never label an unverified regulatory requirement as confirmed. The editable workbook Atlas_Pilar4_Plantillas_EN_v2.xlsx includes a MasterPlan register.

If the change also brings supplier payments forward or delays customer receipts, update the cash-flow and reserves forecast. A decision that fits the overall budget may still fail because cash is unavailable when a payment is due.

For a solo founder, a linked spreadsheet and change log may be sufficient. Teams with contractors, approvals, or legal constraints need stronger version control and accountability. The practical test is whether another person can determine what will be delivered, what remains unknown, and which evidence supports each decision without relying on our memory.

What you should have when you finish

A master index linked to working registers, a bounded delivery objective, a decision log and a review routine. The launch-readiness guide provides the final checks.

Methodological references: [R07], [R01].


Bibliography and references