AI Business Consulting
What is the most important problem that needs to be solved?
Engineers solve problems, consultants sell their services. What bridges the gap?
Most organizations adopting AI are stuck in pilot purgatory. The gap isn't technology — it's tribal imbalance and lack of a problem prioritisation protocol.
What organizations say: "We need AI strategy." What they mean: "We have pilots everywhere and nothing is shipping."
Consulting Psychology
Every major consultancy follows the same meta-pattern: picture the gap, sell the bridge. The sequence is always confusion → clarity → confidence → contract.
| # | Step | What the Consultant Does | Your Version |
|---|---|---|---|
| 1 | Name the pain | Frame the gap. Use data to make it urgent. | "78% adopting AI, most stuck in pilot purgatory" |
| 2 | Diagnose root cause | Show the obvious explanation is wrong | "The gap isn't technology — it's tribal imbalance" |
| 3 | Provide a framework | Make complexity legible. This is where consultants earn. | Three Tribes + Four-Layer Playbook |
| 4 | Prioritize ruthlessly | Decide what to do first. 2x2 matrices, quick wins. | Use Case Prioritization (Layer 3) |
| 5 | Package the engagement | Phases, deliverables, pricing. Low-risk entry, expand on results. | Diagnostic → Pilot → Managed Service |
The consultant's job is to be the person who makes the world legible, then charges to keep it that way. A good story does half the selling. A good presentation does the rest.
The Consulting Checklist
What the best firms actually run through:
- Presenting vs. real problem? Clients describe symptoms, not causes
- Who has power, pain, budget? Sponsor mapping
- Undeniable success in 90 days? Anchor to measurable outcomes
- Simplest explanation for why this hasn't been solved? Structural diagnosis
- 2-3 moves with disproportionate value? Prioritized use cases
- Minimum viable engagement to prove you're right? Paid diagnostic
- How does this expand into recurring revenue? Managed service / retainer
Mapping to Tight Five
The consulting playbook is an external audit of someone else's Tight Five. The sale happens when you show them which P is broken:
| Consulting Step | Tight Five Framework | What the Consultant Does |
|---|---|---|
| Name the Pain | Performance — "How do you know it's working?" | Expose the gap between claimed performance and actual outcomes |
| Diagnose Root Cause | Principles — "What truths guide you?" | Show decisions feel random because there's no shared first-truth |
| Provide a Framework | Platform — "What do you control?" | Assess assets, data, tools, architecture. Show what's missing structurally. |
| Prioritize & Coordinate | Protocols — "How do you coordinate?" | Install the missing operating system. Why success doesn't compound. |
| Package the Engagement | Players — "Who creates harmony?" | Map the Three Tribes. Show the tribal mix is unbalanced. Position yourself as the missing player. |
The deep insight: consultants sell the binding, not the elements. A client usually has all five Ps in some form. What they're missing is the tightness — the coordination that makes five compound instead of compete.
Recursive Application
The Tight Five works at every level of the engagement:
| Level | The Five | Maps To |
|---|---|---|
| Discovery | Capture, Inquire, Design, Platform, People | Diagnostic phase — help them find their five |
| Strategic | Principles, Performance, Platform, Protocols, Players | Roadmap deliverable — help them define their five |
| Tactical | 5 Facts → 5 Questions → 5 Answers → 5 Ideas → 1 Decision | Each use case one-pager — help them act on their five |
The diagnostic demonstrates the framework's value. You use the Tight Five to assess the client, and the assessment is the sale — once they see through the lens, they can't unsee which P is broken.
Four-Layer Playbook
Layer 1: Business Alignment
Anchor every discussion in business value, not technology.
| Area | Key Questions |
|---|---|
| Objectives | What 2-3 business metrics should AI move first? |
| Scope | Which units/processes are in-scope in the next 90 days? |
| Sponsor | Who signs off, who blocks, who operates day-to-day? |
| Constraints | What data/regs/brand rules can't be violated? |
| Success | What proves this was worth it within a quarter? |
Layer 2: Readiness Diagnostic
Run in a 2-4 week diagnostic phase.
| Dimension | What to Assess |
|---|---|
| Strategy & Leadership | Named AI owner? Documented vision linked to budget? Clear principles on where AI will/won't be used? |
| Data Foundation | Accessible, permissioned data? Governance policies? Sensitivity classification? |
| Technology & Architecture | Approved platforms? API/integration patterns? MLOps capability? |
| Governance & Risk | AI risk policy? High-risk review process? Output monitoring? |
| People & Skills | Internal AI training? Champions in teams? Capacity without derailing ops? |
Layer 3: Use Case Prioritization
| Criterion | Question |
|---|---|
| Business Impact | Quantifiable upside in time, cost, revenue, or risk reduction? |
| Feasibility | Data availability, technical complexity, dependencies? |
| Time-to-Value | Result inside 4-12 weeks with a scoped pilot? |
| Adoption Likelihood | Clear owner, motivated team, workflow people want to improve? |
Layer 4: Delivery & Governance
| Phase | Duration | Purpose |
|---|---|---|
| Phase 0 - Diagnostic | 2-4 weeks | Readiness checklist, discover/prioritize use cases, strategy + roadmap |
| Phase 1 - Pilots | 6-12 weeks | Implement 1-3 high-ROI use cases with clear metrics and governance |
| Phase 2 - Scale | Ongoing | Expand successful pilots, build reusable agents, train teams, formalize operating model |
The Three Tribes
AI transformations fail when tribes are unbalanced.
| Tribe | Question They Ask | What They Provide |
|---|---|---|
| Explorers | "What if we tried...?" | Discover options, frontier awareness |
| Automators | "How do we operationalize this?" | Scale validated ideas, integrate systems |
| Validators | "How do we ensure quality and safety?" | Standards, compliance, trust |
The failure pattern: Explorers generate pilots → Automators can't operationalize → Validators block everything → Pilot purgatory.
The solution: Tight fives that pass through all three tribes before shipping.
The Three Tribes model is the Players P applied recursively — Explorers, Automators, and Validators need their own tight five to ship anything, which is exactly the "pilot purgatory" failure pattern.
Pricing Model
| Tier | Deliverable | Price Range |
|---|---|---|
| Diagnostic | Readiness baseline + prioritized use case roadmap | $5K-15K |
| Pilot | 1-3 implemented use cases with metrics and governance | $15K-50K |
| Managed Service | Ongoing AI worker operations + optimization | $3K-10K/month |
Providers to Study
Context
- Tight Five Framework — The diagnostic lens consultants sell
- Trades Consulting — Same playbook, different industry
- Pitch Decks — Packaging the engagement for sale
- Presenting — Delivering the pitch
- Storytelling — The narrative that sells
- Pictures — Picture the gap, sell the bridge
- Cost Centres — What you're helping them fix
- Solopreneur Model — Alternative delivery model
Questions
What is the most important question this topic raises that current discourse tends to avoid or understate?
- Which assumption in the standard framing of this topic is most likely to be wrong in a 5-year horizon?
- How does the DePIN or agent-native lens change what matters most about this topic?
- Which first principle, if violated, would make the analysis of this topic fundamentally incorrect?