/02Whole Business Map
A one-page way to map departments, workflows, AI jobs, owners, proof, and rollout order without making fifty slides.
Leader
12 min
Open +
Whole Business Map
A one-page way to map departments, workflows, AI jobs, owners, proof, and rollout order without making fifty slides.
Leader
12 min
A useful AI framework does not start with models. It starts with the work your business already does and makes the next change visible.
Leave knowing / Brief the whole business on one page
The rule
Do not make an AI strategy by listing tools. Make it by listing work. Every department is a collection of workflows, and every workflow has steps, owners, inputs, decisions, outputs, and evidence.
AI belongs where one of those steps can become faster, clearer, cheaper, more consistent, or newly possible without creating more risk than value.
Department -> workflow -> AI job -> human owner -> evidence. That is the map.
The board view
Ten departments. One question each.
These are examples, not prescriptions. Map the workflow before choosing the technology.
| Department | Good first question | Possible AI job |
|---|---|---|
| Sales & Revenue | Where do reps lose time before the next customer action? | Research, call prep, CRM cleanup, follow-up drafts |
| Marketing & Growth | Which content or analysis cycle repeats every week? | Briefs, variants, insight synthesis, reporting |
| Customer Success | Which questions arrive again and again? | Case summaries, answer drafts, risk signals |
| Business Operations | Where do people copy information between systems? | Intake, routing, document handling, exception triage |
| People & Culture | Which admin step can be assisted without outsourcing judgment? | Onboarding, policy Q&A, role materials |
| Data & Analytics | Where is analyst time spent preparing instead of interpreting? | Query assistance, data documentation, recurring summaries |
| Finance | Which recurring review starts with gathering and explaining? | Variance explanations, document extraction, scenario drafts |
| Product | Where does customer evidence become slow to synthesize? | Feedback clustering, research synthesis, spec first drafts |
| Engineering | Which low-risk development loops are repetitive? | Tests, documentation, code review assistance, migrations |
| Design & UX | Where can exploration happen faster without replacing taste? | Research synthesis, content variants, prototype assistance |
Rollout order
The order matters more than the logo on the tool.
1. Assist one person.
Start with a reversible task where the user can see and judge every output.
Output / Assist
2. Improve one workflow.
Connect the AI to the context and rules needed to remove a real bottleneck.
Output / Workflow
3. Connect the department.
Standardize access, ownership, measurement, and escalation across related workflows.
Output / System
4. Add bounded agents.
Only after permissions, tools, recovery, and proof exist should software act through multiple steps.
Output / Agent
The one-page brief
Put these six fields beside every AI idea.
Outcome: what gets better for the business or customer?
Workflow: what exact work changes?
AI job: summarize, draft, classify, retrieve, decide, or act?
Owner: which human is accountable for the workflow?
Boundary: what is the AI not allowed to do?
Proof: what number or evidence will tell us it worked?
Build your map
Turn an org chart into a working AI map.
Try the template
Act as an AI transformation architect. I will describe my business and departments. For each department: 1. List the 3-5 most important recurring workflows 2. Identify the best AI assistance opportunity in each workflow 3. Name what should remain human 4. Name the data or tools AI would need 5. Suggest one measurable proof of value 6. Rate implementation difficulty: Low / Medium / High Then recommend the first 3 pilots for the whole business. Prioritize useful, reversible work over flashy demos. Business: [DESCRIBE BUSINESS AND DEPARTMENTS]
Notes / Subscribe
Get the next note.
No schedule. No content calendar. A note goes out only when there is something worth understanding.