Consulting Thinking Capability System

CONSULTING THINKING CAPABILITY SYSTEM

From meeting notes to consulting judgment

Book I · how a consultant thinks · Filling boxes isn't thinking — the value is in the "diagonal" (connecting points across different subjects and times into one recommendation). Collapse the sidebar with ☰.

Where should I start?

What this is · AI-Native eBook

A traditional book is read; this one is used — paste your next real meeting into the tool and get scored, find the missing diagonal, and generate a summary draft on the spot.

North star: "I don't want you to remember what the client said — I want you to judge what the client truly needs, and how we win the next step."

Honest positioning: not a "new strategy framework," but an operationalized system that makes consulting thinking trainable, scorable, and verifiable — standing on Ohmae's 3C, TOWS, and Temporal Competitive Analysis, filling the missing "subject × time" orthogonal grid between them.

Theory · The Matrix (Input vs Output)

📥 Customer input = raw material you hear (given to you)　🧠 Consultant output = what you produce (where the fee is)

Lens ＼ Time	Past · Cause	Present · Need	Future · Solution
Customer	📥 pain / legacy 🧠 root cause: you infer	📥 surface need + 🧠 true pain	🧠 rollout path + unspoken need ★
Competitor	🧠 how they built advantage	🧠 what they offer / blind spot	🧠 how they evolve / new moat
Company (us)	🧠 how our capability formed	🧠 can / cannot solve	🧠 capability to grow / how to earn
Place (env.)	🧠 structural shift happened	🧠 regulation / tech tailwind	🧠 irreversible trend / window

🔑 Only the top-left two cells (customer × past/present surface) are 📥 input; the other 7 cells are mostly 🧠 output.
A C1 person can only fill the 📥 cells = copying the client. The consulting value is entirely in the 🧠 output cells + the diagonal. 📥 is given to you; 🧠 is your value.

Three readings · The diagonal is the core

Vertical = vector: past→present→future; derive the future from "what force pushed past into present" (don't guess).
Horizontal = gap: same time, customer vs competitor vs us vs environment; asymmetry = opportunity.
Diagonal = the fee ★: connect across subject × across time. Typical: competitor's present blind spot × customer's future need × environment's past shift → recommendation.

🛑 Filling ≠ thinking. Nine boxes filled but no diagonal is worse than "only the customer column, but one sharp diagonal."

Capability ladder C1–C5

C1 Record→C2 Organize→C3 Analyze→C4 Plan (one diagonal)→C5 Advise (multiple diagonals + unspoken + us/env.)

Most people are stuck at C1. This is the ladder across the C1→C5 gap.

Beginner: do only 5 steps the first time

Write the customer's exact words
Find the true pain (surface A → root B)
Derive what happens in the future (+ the unspoken)
Add one competitor or environment factor
Write one diagonal recommendation

You don't need to memorize prompts or the grid — the tool below guides you.

🛠 Tool: Consulting Meeting-Summary Workbench

🔒 When notes contain client confidential content, use a company-approved internal AI or private endpoint — do not paste to a public cloud.

1. Customer's exact words (facts) 📥 input 2. True pain (surface → root) 3. Present need (job to be done) 4. Future recommendation + unspoken need ★ 5. Competition & environment (blind spot / trend) 6. How we (TigerAI) enter (solve / can't / which standard product) 7. Next step (who / when / what) ⭐ Diagonal (use this pattern)

🚀 Send straight to the AI coach (optional; needs an n8n Webhook URL from your manager)

If set, the 7 parts + diagonal are POSTed as JSON to n8n; the AI reply shows below. The URL is saved locally. (n8n must allow CORS.)

7 pre-submit questions

Any "No" → not ready. Watch Q2, Q3, Q6.

🎯 Self-Scoring Rubric (managers too)

Six items, 0–2 each. Rule: no diagonal → capped at C3; only answering what was asked → max C4.

Has one defensible diagonal Caught an unspoken need

/ 12

📋 Copy to AI for gatekeeping (uses a Prompt, not built-in AI)

🛑 If your report has no sentence shaped like "Because the competitor now…, plus the environment trend…, the customer will in future…", do not hit submit.

This page does not call AI itself. Copy your notes to the Prompts on tab ④ or to a Slack bot; the AI returns C-level, the missing diagonal, and a summary draft.

🧑‍🏫 Manager Socratic questions (ask, don't scold)

"In three years, when they scale, what does this pain become? What should we propose now?" (force future)
"How do competitors solve this? Does their product do it now? Could they overtake?" (force competitor)
"If the client expands to another city/country, how do regulation/tech/competition differ?" (force environment)
"What can we solve / not solve here? How do we price? What must be productized?" (force us)
"Connect those three points into one recommendation — why does this line hold?" (force diagonal)

Training order: customer column + diagonal first → add competitor → add environment/us. Depth before breadth.

Use the AI coach: just talk

With the consulting-thinking-coach skill (Claude Code / Codex / Antigravity), just say:

What you want	Say to the AI
Notes → consulting summary	"Critique these meeting notes with the consulting thinking matrix"
Score & gaps	"Score this with the rubric — what C level am I?"
Connect a diagonal	"Which diagonal is missing? Connect one for me"
(Manager) Socratic	"Generate 4 questions pushing future/competitor/environment"
Client-facing insight	"Turn into a client insight memo, hide internal reasoning"

⚡ n8n: zero-friction adoption (Slack/Line bot)

After a meeting, drop notes into #consulting-coach or a Line bot → n8n auto-returns C1-C5 score + missing diagonal + summary draft.

[① Trigger Slack/Line] → [② IF filter] → [③ Set notes] → [④ AI node: structured JSON] → [⑤ Set format] → [⑥ Reply] → [⑦ Log→Sheet (auto-collect scores)]

④ AI-node output schema (for ⑤ formatting / ⑦ logging):

{ "level": "C3", "score": 8, "missing_diagonal": true, "diagonal": "competitor now… × customer future… → recommend…", "summary": "six-part consulting summary…", "next_action": "who / when / what" }

🔒 Step ⑦ grows your empirical dataset. When notes contain client confidential content, point the LLM node to a company-approved internal/private endpoint — not a public cloud.

📋 Copy-ready prompts (paste into any AI)

P1 · Notes → consulting summary + score

You are a Consulting Thinking Capability System coach. Rewrite the meeting below into a consulting summary (six parts: facts / true pain / present need / future + unspoken / competition & environment / how we enter) + next step, and explicitly write one diagonal (competitor or environment at one time × customer at another time → recommendation). Then score with the rubric: six items 0-2 (no diagonal caps C3; only answering what was asked caps C4); give total, C-level, and the most important missing cell. Separate signal vs inference; mark anything unverified as TBD; do not fabricate. [Paste meeting notes]:

P2 · Score only

Score this summary with the capability rubric: six items 0-2 each with reasons; no diagonal caps C3, only answering what was asked caps C4; compute total (12), classify C1-C5, name the most important missing cell and one concrete step to move up a level. [Paste summary]:

P3 · Find the missing diagonal

Map this against the 3-subject (customer/competitor/us) + environment × 3-time matrix; point out which cells are empty and which diagonal is unconnected. Offer 1-2 plausible diagonal candidates for me to verify — do not assert anything without evidence. [Paste summary or notes]:

Which tools work

Claude Code / Codex: skill built (consulting-thinking-coach); plain language triggers it.
Antigravity (AG): same skill mechanism.
n8n: build it into a Slack/Line auto-coach bot.
Any AI chat box: paste P1–P3 above; works without the skill installed.

Author: Morris (Yeh-Hsing) Lu · LinkedIn · ORCID 0009-0006-5373-0586
Consulting Thinking Capability System · single-file interactive book (offline-ready, Ctrl+P to print) · Apache License 2.0