30-Session Programme · 25 Minutes Per Day · Free · For leaders in Nigeria’s oil and gas sector
WEEK 6 OF 6 · THE LEADERSHIP MANDATE · SESSIONS 26–30 · WORK AT YOUR OWN PACE
Part Six · The Leadership MandateSESSION 26
Building Capability That Stays
Why This Matters
In Session 25 you mapped the capability each role in your company needs. A map is not yet a workforce. Today you turn it into something the board can fund and hold you to: a 90-day capability pilot and a scoreboard that measures capability built, models owned and dependency reduced, each against a baseline. The uncomfortable part comes first. Ask what your company paid outside parties last year for work its own people could have done, and which models it relies on that nobody inside can maintain. Chapter 16 is plain about where the risk lies: not in competitors buying newer tools, but in competitors owning the capability while you keep renting it.
25-Minute Module
0:00 – 5:00
The value map: what renting costs you
- Map the value of capability before investment begins. Start with avoidable cost and dependency. What did your company spend last year on external vendors for work its own people could do? Which decisions were delayed because no one internal could interpret the model? Which models cannot be maintained without a foreign consultant?
- Then the other three parts of the book's value map: performance, because capability is the multiplier on every other digital investment; sovereignty and resilience, because a company that can explain and defend its own models is more trusted than one that rents them; and scalability, because capability compounds.
- THE DOCTRINE
The risk of inaction is not that competitors will buy newer technology. The real risk is that competitors will own the capability to use it while others keep renting it.
- Inaction also has a compliance edge. As NUPRC moves toward measurement-based methane and GHG reporting, operators need people who can produce credible, traceable data and govern the models behind it. Capability, the book says, is becoming a compliance requirement, not just a performance advantage.
If you cannot say what your company paid others last year to think for it, that is the first number to find.
5:00 – 11:00
The 90-day capability pilot
- Do not try to upskill the whole company. Target one role group, asset team or capability gap where the pain is clear: field analytics for one asset's engineering team, say, or data engineering for the function that most often stalls AI work.
Days Purpose What it produces 1–30 Discovery and baseline: identify the gap, quantify its cost, map the target roles, define ownership A charter: scope, baseline, target capability levels, success metrics, stakeholders, governance and a link to a real operating decision 31–60 Build in context: role-based curriculum, hands-on training on the company's own data and tools, mentoring, applied projects People demonstrably doing something they could not do before, on a real asset 61–90 Applied adoption: live work, reviewed in normal operating routines A record of which capabilities changed real decisions and produced measurable value - Before any course, state the value case in dependency reduced and performance improved, validate it with finance, and design how the skill will be used: which decision, which routine, which mentor. End with a disciplined decision: continue, expand, adjust or stop.
- THE DOCTRINE
A pilot is successful only when it proves a repeatable pattern: that capability built in context changes decisions and creates value. Anything less is a course, not a transformation.
Choose the gap that already costs you money. A pilot aimed at goodwill will be judged on goodwill.
11:00 – 17:00
A capable human in the loop
- AI observes, detects, ranks, recommends and tracks; humans validate, decide, approve and execute. That human-in-the-loop model works only if the human is capable.
- THE DOCTRINE
A validation step performed by someone who cannot interpret the model is not a safeguard. It is a rubber stamp.
- Governance is itself a capability. The book makes the NIST AI Risk Management Framework and ISO/IEC 42001 the syllabus for governance training, and cites NIST SP 800-82 for operational technology security: a workforce that cannot secure its own systems has not achieved ownership.
- IN THE FIELD · THE WORKED CASE
- Field. A supervisor at a fictional marginal field operator signs off remaining-useful-life estimates from a vendor-hosted pump model without being able to say what data it used. The capability pilot is field analytics for that asset's team, so the sign-off becomes a real validation.
IN THE BOARDROOM · THE PARALLEL- Boardroom. The board asks of the same model what data it used, how current it is, whether the company can maintain it and who is accountable when it is wrong. If any answer is to phone the vendor, the company is renting its intelligence layer.
- Ownership is not zero partnership. The book's Local Content 2.0 trajectory, which it labels illustrative, is a phased shift from imported AI toward indigenous ownership, with NCDMB's capacity-development mandate as its institutional home.
Ask who in your company could explain your most important model to a regulator without a vendor on the call.
17:00 – 22:00
Scaling, rhythm and the scoreboard
- Scale rests on three principles: standardisation of capability levels and assessment; modularity, reusable modules combined into role pathways; and ownership, an executive sponsor, role owner and performance outcome for every track. Then put capability into the rhythm: can the team validate today's alerts, which gap caused this week's avoidable vendor call, how is dependency trending this month?
- THE DOCTRINE
Executives should not measure capability by courses run, people trained or certificates issued. Those are activity metrics. The board should measure capability by the ownership and performance it produces.
Indicator What it tells the board AI literacy completion Share of each role at its target literacy level Local model ownership Share of models your own people build, maintain and govern Workflow adoption Whether the tools are used in daily work Field reporting quality Quality, completeness and timeliness of captured data Local vendor capability Whether Nigerian vendors can deliver AI-enabled services Training-to-performance conversion Rate at which training becomes measurable improvement Reduced external dependency Trend in vendor spend on work you could do yourself Digital career pathways Whether pathways exist and retain talent - Each needs a baseline. Low adoption despite high completion means the application loop, not the curriculum, needs attention. Add the book's final metric, learning velocity. The book's scoreboard figure is illustrative; your baselines must be your own.
If two of the eight have no baseline, say so. An honest blank is worth more than an invented figure.
22:00 – 25:00
Before you close the laptop
- Write down the capability gap that costs your company most, and the one number that would prove it.
- Nigeria's 3 million barrel per day ambition, the book says, requires deeper technical capacity, and that capacity is, in the end, people.
- This session produces Section 26 of your 90-Day AI Survival Game Plan: your capability pilot and scoreboard.
Choose ownership over convenience, and write the pilot so that someone in your company can be held to it.
Key Terms
Capability pilot
A 90-day pilot on one role group, asset team or capability gap, with a charter, baseline and senior sponsor, ending in a stop-or-scale decision.
Local model ownership
The share of the models the company depends on that its own people build, maintain and govern.
Training-to-performance conversion
The rate at which training becomes measurable performance improvement.
Learning velocity
How quickly lessons from applied projects, successful and failed, feed back into how people are developed.
Self-Check — Before You Submit
1. In a fictional EPC company, HR reports that nearly every engineer has completed an AI course and proposes rolling it out company-wide. Workflow adoption remains low. What do you do?
- A. Approve the rollout, since completion is high
- B. Fix the application loop first, deciding which decision each skill improves and in which routine, and scale only once adoption is proven
- C. Replace the course with a more advanced one
- D. Stop measuring adoption, since HR cannot control it
B. Fix the application loop first, deciding which decision each skill improves and in which routine, and scale only once adoption is proven — Low adoption despite high completion means the application loop needs attention, and scale rewards proven application. Approving the rollout scales unapplied training, which the book says multiplies cost.
2. According to Chapter 16, what makes human-in-the-loop genuinely safe?
- A. A senior signature on every model output
- B. A vendor warranty on the model
- C. Keeping AI out of field decisions
- D. A human capable of interpreting, verifying and governing the model
D. A human capable of interpreting, verifying and governing the model — The loop only functions if the human is capable; otherwise validation is a rubber stamp. A senior signature without that capability is exactly the rubber stamp the book warns about.
3. Which are the book's three principles for scaling a capability programme?
- A. Standardisation, modularity and ownership
- B. Literacy, application and ownership
- C. Speed, scale and certification
- D. Centralisation, outsourcing and audit
A. Standardisation, modularity and ownership — Chapter 16 names standardisation, modularity and ownership. Literacy, application and ownership are stages of the maturity curve, a different idea.
Today’s Assignment
Write Section 26 of your plan: your company's 90-day capability pilot and capability scoreboard, each measure with a baseline.
- Name the capability gap that costs your company most and quantify it: vendor spend on work your people could do, decisions delayed, models nobody inside can maintain. Where you lack a figure, name who will find it by day 30.
- Write the charter for one role group, asset team or gap: scope, baseline, target capability levels, success metrics, stakeholders, governance and the operating decision it improves.
- Lay out days 1–30, 31–60 and 61–90 with the sponsor and role owner by role, and the routine and mentor for applying the skill.
- Build the scoreboard: the eight indicators, each with today's baseline or an honest blank and an owner, plus learning velocity.
- Say honestly where your company counts activity instead of capability, and close with your continue, expand, adjust or stop criteria. If your company has no AI in use yet, work from the vendor dependency or decision in front of you.
YOUR 90-DAY PLAN
Section 26 — Our capability pilot and scoreboard.
How this is marked — the rubric
| Names the costliest capability gap and quantifies it, or names who will by day 30 | 20 |
| A charter with scope, baseline, target levels, success metrics, governance and one operating decision | 25 |
| Three phases with sponsor and role owner by role and a designed application loop | 20 |
| All eight indicators with baselines or honest blanks and owners, plus learning velocity | 25 |
| An honest account of activity metrics, with stop-or-scale criteria | 10 |
| Total · pass mark 70 | 100 |
Submitting session 26
Write your answer in Word or Google Docs. The first line must read
Write your answer in Word or Google Docs. The first line must read
Session: 26, or the submission is refused. Upload the Word file (.docx), or switch to the Google Doc tab and paste your document’s share link.Marked against the rubric above and returned as a written letter, saved to your progress page. Do not submit until you have passed the self-check above.Capabilities Covered
Capability value mappingCapability pilot designHuman-in-the-loop competenceCapability scoreboardScaling without losing control
Part Six · The Leadership MandateSESSION 27
The 12-Pillar AI Game Plan
Why This Matters
The wrong question in this market is which AI tool to buy. The right one, Chapter 17 says, is what your company must become so that AI can actually improve performance. Today you meet the book's answer: twelve pillars, from executive AI fluency to ecosystem partnerships, around which every Nigerian oil and gas company should build its roadmap. They are not a menu. They are sequenced, with governance and data first and autonomy last, and the book calls skipping the sequence the most common and most expensive failure mode in this industry. You will score your company against all twelve, honestly, and choose the three that come first.
25-Minute Module
0:00 – 5:00
What the company must become
- AI is not a product you buy; it is a property of how the company operates. Bolt the smartest tool onto a siloed, low-trust operating model and you get faster confusion, not better barrels. The book's word is survival because AI compounds a decision-quality advantage between intelligent and reactive operators, slowly at first, then beyond catching. NNPC's and NUPRC's production targets, the book adds, will not be delivered by drilling alone.
- THE DOCTRINE
The wrong question, which AI tool should we buy, anchors leadership in vendor hype. The right question, what must our company become so AI can actually improve performance, anchors it in operating capability.
- Readiness reduces to six shifts: manual reporting to real-time visibility; reactive maintenance to predictive reliability; fragmented data to governed intelligence; isolated departments to integrated workflows; vendor dependence to local digital capability; executive opinion to evidence-based decision-making. None needs exotic technology. All need management courage.
When a vendor opens with the tool, ask which of the six shifts it serves.
5:00 – 12:00
The foundations and the reliability core: pillars 1 to 6
- The foundational pillars (fluency, data, visibility, governance) make the high-value pillars (optimisation, twins, surveillance, autonomy) possible and safe: the Sequencing Discipline applied to the whole company.
Pillar What the book directs What good looks like 1. Executive AI Fluency Board-level AI literacy as the first project; strategy owned at the executive table, not by IT or a vendor A fluent CEO can ask for the model inventory, who owns this data and where a reported number came from, and get real answers 2. AI-Ready Data Foundation A data contract per critical domain; build around the highest-value decisions, not an enterprise data lake Every critical domain has a contract in force 3. Operational Visibility One operational picture; the daily review run on exceptions You know your decision latency, signal to validated action, without a month-end report 4. Predictive Maintenance and Asset Reliability Start with pumps, compressors and critical rotating equipment; close the loop to procurement Alerts are failure-mode specific and reach a named owner 5. Production Optimisation Decline-curve analytics on every well plus constraint management; deferment causes and underperforming wells first A daily rhythm finds this week's binding constraint, then the next 6. Digital Twins and Living Asset Intelligence Lean, fit-for-purpose twins; earn the descriptive stage first The twin stays synchronised and is tied to real decisions such as workover sequencing - The book names thirteen critical data domains, from asset and production to workforce and client or tender. On brownfield assets the first work is data archaeology. And the investment pays twice: the lineage that makes AI trustworthy is what NUPRC's move to measurement-based MRV reporting demands.
- THE DOCTRINE
An alert that reaches no named owner is operational noise, and an inventory of alerts nobody acts on is worse than no system at all.
Score pillar 1 first. If the executive table cannot pass the fluency test, every other score is optimistic.
12:00 – 18:00
Commercial, security, compliance and people: pillars 7 to 12
Pillar What the book directs What good looks like 7. AI-Enabled Tendering and Commercial Excellence A retrieval-based memory of past bids; automated compliance and Nigerian Content matrices; bid/no-bid discipline Proposal effort goes only to winnable, profitable work 8. Pipeline Surveillance, Crude Theft, and Asset Security Sensor fusion; route-risk scoring for drones, patrols and response teams Crude-loss reduction and detection-to-response time are board metrics with baselines 9. AI-Enabled Compliance and MRV Lineage and evidence management built into the reporting workflow Every regulatory number can be reconstructed on demand 10. Cybersecurity and Operational Trust A safe, governed AI environment, not a blanket ban; classify data, lock the OT boundary, log everything No uncontrolled bridge between enterprise IT and field OT 11. Local Content 2.0 and Workforce Reskilling Capability mapped to roles; data-engineering and governance tracks first Data-engineering and AI-governance capability is built and owned in-house, and every role has the right depth 12. AI Partnerships and Ecosystem Building Partners chosen for implementation discipline, not demo polish They can take you from a slide to a number - On pillar 9: NUPRC has directed upstream operators toward standardised templates and measurement-based methane and greenhouse-gas reporting, and its guidelines apply to operators and service providers for Monitoring, Reporting and Verification. An untraceable figure is a regulatory finding waiting to happen.
- IN THE FIELD · THE WORKED CASE
- Field. At a fictional drilling and well services company, a tender engineer pastes a client's well documents into a public AI tool. The book names this everyday leakage as one of the most underestimated exposures in the industry.
IN THE BOARDROOM · THE PARALLEL- Boardroom. The tempting response is a blanket ban. The book's direction is a governed AI environment, yes within controls, because a ban drives shadow usage into public tools.
Pillar 12's point is that the binding scarcity is integrative leadership, not technology. Ask whether your company has it.
18:00 – 22:00
Scoring honestly and choosing three
- THE DOCTRINE
The twelve pillars are not a menu; they are an interlocking system with a deliberate order. Governance and data first, autonomy last.
- Score each pillar against what good looks like in the tables above: in place, partial or absent, with one line of evidence you could show a sceptical director tomorrow. For pillar 1 the book gives the test itself: the three questions a fluent CEO can ask.
- Rank the three that come first using two filters. The Sequencing Discipline: a high-value pillar ranked above a missing foundation needs a written reason. The Operations-Excellence Lens: each of the three must connect to the one loss you named in Section 1.
- If none of pillars 1 to 3 is among your three, check your scores again. Attempt the exciting pillars without the foundations, the book warns, and the programme will stall, leak money or create risk faster than value.
Score the company you run, not the one in the annual report.
22:00 – 25:00
Before you close the laptop
- Write down your three pillars, in order, and the loss each serves.
- This session produces Section 27 of your 90-Day AI Survival Game Plan: your company scored on the twelve pillars, and the three that come first.
- Next session you see where the book says your kind of company should start. Check then whether your three agree.
An honest absent is the most useful score on the page. It tells you where the ninety days begin.
Key Terms
Data contract
For each critical data domain: a named owner, a standard definition, a validation rule, a quality threshold and a reporting standard.
Decision latency
The elapsed time from signal to validated action; the variable the book says everything else in operational visibility rides on.
Data archaeology
On brownfield assets, reconstructing an accurate record of what actually exists before any model is trusted on it, because documentation often does not match the asset.
Route-risk scoring
Ranking pipeline routes by exposure so scarce drones, patrols and rapid-response teams concentrate where risk is highest.
Self-Check — Before You Submit
1. A fictional upstream operator's board wants to start with an AI agent that adjusts production settings. Its production data has no owners or definitions. What do you recommend?
- A. Proceed, because high-value pillars pay most
- B. Build an enterprise data lake first
- C. Start with fluency, data contracts and visibility for the decisions that matter, then sequence the rest
- D. Buy the agent and fix data as problems appear
C. Start with fluency, data contracts and visibility for the decisions that matter, then sequence the rest — The book puts governance and data first and autonomy last, and calls skipping the sequence the most expensive failure. The enterprise data lake is what the book tells you to resist.
2. Which elements make up a data contract under pillar 2?
- A. A named owner, standard definition, validation rule, quality threshold and reporting standard
- B. A vendor, a licence fee, a service level and a renewal date
- C. A dashboard, a data lake and a model
- D. A regulator, a template and a deadline
A. A named owner, standard definition, validation rule, quality threshold and reporting standard — Those are the five elements of the book's data contract. A vendor agreement is commercial and says nothing about who owns the data or how quality is judged.
3. Staff at a fictional fabrication yard paste client drawings into public AI tools. What does pillar 10 direct?
- A. Ban all AI use
- B. Wait for an incident
- C. Have staff sign a policy
- D. Provide a governed AI environment, yes within controls, and classify the data
D. Provide a governed AI environment, yes within controls, and classify the data — A ban is tempting, but the book says it drives shadow usage into public tools, leaving the leakage risk in place and out of sight.
Today’s Assignment
Write Section 27 of your plan: your company scored honestly against the twelve pillars, and the three that come first.
- Score all twelve pillars as in place, partial or absent against what good looks like in this session's tables, with one line of evidence each.
- Name at least two pillars where your company has told itself a better story than the evidence supports.
- Rank the three pillars that come first, showing how each connects to your Section 1 loss and where it sits in the sequence. Justify any high-value pillar ranked above a missing foundation.
- For each of the three, name the owner by role and the first move from the book's direction.
- Name one pillar you are deliberately deferring, and why. If your company has no AI in use yet, score it anyway: most pillars test operating discipline, not software.
YOUR 90-DAY PLAN
Section 27 — Our company scored on the twelve pillars, and the three that come first.
How this is marked — the rubric
| All twelve pillars scored against the session's standard, with evidence for each | 30 |
| At least two over-generous self-assessments corrected honestly | 20 |
| Three pillars ranked with explicit links to the Section 1 loss and the sequence | 30 |
| Owner by role and first move for each of the three, plus a reasoned deferral | 20 |
| Total · pass mark 70 | 100 |
Submitting session 27
Write your answer in Word or Google Docs. The first line must read
Write your answer in Word or Google Docs. The first line must read
Session: 27, or the submission is refused. Upload the Word file (.docx), or switch to the Google Doc tab and paste your document’s share link.Marked against the rubric above and returned as a written letter, saved to your progress page. Do not submit until you have passed the self-check above.Capabilities Covered
Operating-model readinessTwelve-pillar assessmentSequencing the roadmapData contractsShadow AI risk
Part Six · The Leadership MandateSESSION 28
The Game Plan for Your Kind of Company
Why This Matters
The twelve pillars apply to every company in the industry, but not in the same order. Chapter 17 sets out where each type of company should start: upstream operators, marginal field operators, EPC and construction companies, drilling and well services, marine and logistics, fabrication yards, and downstream, midstream and gas companies. Today you take the plan for your type, test it against the three pillars you ranked in Section 27, and choose the KPIs your board will track, each with a baseline. Then you hold up the book's eight-part leadership model and ask the harder question: which of those capabilities you are currently missing. Pop Quiz 6 opens at the end.
25-Minute Module
0:00 – 4:00
Same pillars, different starting points
- Every segment draws from the same twelve pillars but starts in a different place, and each type's list in the book comes with one objective. Read the objective as closely as the list: it tells you what the board should see move.
- THE DOCTRINE
Focus is the discipline: start where the value is clearest for your specific business model, prove it, and expand.
- Use the company type you named in Section 1. If your company straddles two, choose the one where your costliest loss sits, and say so in the plan.
Read your own type's row twice and the others once. The others describe your clients, contractors and competitors.
4:00 – 11:00
The plan by company type
- The book's priorities and objective for each type:
Company type Priorities the book names Objective Upstream operators Production optimisation, predictive maintenance, digital twins, pipeline surveillance, emissions reporting, asset-integrity analytics, workover prioritisation, field-operations dashboards, contractor-performance analytics Protect production, reduce deferment, improve reliability and increase asset visibility, turning the existing asset base into the source of the barrels the national targets require Marginal field operators Low-cost production analytics, maintenance prioritisation, cash-sensitive operating dashboards, workover decision support, lean digital-twin models, decline monitoring, contractor-cost analytics, compliance automation Operate with major-level intelligence without major-level overhead, using AI as a force multiplier on thin margins where candidate selection and deferment control decide survival EPC and construction AI-assisted project controls, schedule-risk prediction, cost-variance detection, procurement-expediting intelligence, subcontractor scorecards, HSE-risk prediction, document-control automation, tender intelligence Deliver faster with stronger cost control and fewer execution surprises, because the unforecast surprise is where margin disappears Drilling and well services Rig-equipment reliability, crew-competency tracking, non-productive-time analysis, maintenance forecasting, spare-parts planning, HSE-observation analytics, well-intervention intelligence, client-reporting automation Reduce non-productive time and improve service reliability, the two numbers that drive client retention and the company's own economics Marine, logistics and vessel operators Vessel-maintenance prediction, route optimisation, fuel-consumption analytics, crew-scheduling intelligence, port-delay tracking, contract-performance monitoring, asset-utilisation dashboards, safety-incident prediction Improve utilisation, reduce downtime and protect contract performance Fabrication yards and workshops Equipment-utilisation analytics, welding-productivity tracking, quality-inspection intelligence, material traceability, workforce-productivity analytics, project-progress verification, inventory optimisation, HSE-trend detection Convert yard activity into measurable productivity and quality Downstream, midstream and gas Demand forecasting, inventory optimisation, trucking-and-logistics intelligence, facility reliability, gas-pipeline monitoring, customer-and-contract analytics, safety-and-compliance reporting, energy-consumption optimisation Improve supply reliability, commercial visibility and operational efficiency, increasingly central as Nigeria commercialises gas - Compare the list for your type with your three pillars from Section 27. Where they agree, you have a starting point. Where they disagree, one of them is wrong for your company, and you should be able to say which and why.
Where the book's list for your type and your own ranking disagree, write the disagreement down. It is usually where the plan becomes honest.
11:00 – 17:00
The KPIs every company should track
- AI must be measured or it cannot be managed, and against a baseline, because without a defined starting point no improvement can be credibly claimed. The book's set is sixteen KPIs across four themes.
Theme KPIs Reliability and operations Downtime reduction; maintenance compliance; equipment availability (and, underneath them, mean time between failures and mean time to repair); production-deferment reduction; field-report timeliness; procurement-delay reduction Commercial Bid-submission quality; tender win-rate improvement; contractor-performance score; cost savings from AI use cases; revenue protected or recovered Compliance and data Compliance-submission accuracy; data-quality score; safety-observation closure rate Capability AI adoption rate; workforce-training completion - Count it: six, five, three and two make sixteen. MTBF and MTTR sit underneath the reliability measures, in the book's words, rather than as separate lines.
- THE DOCTRINE
Activity metrics such as models deployed, dashboards built and people trained are not value. Outcome metrics moved against a baseline are.
- The book asks for a focused set across the four themes, so the whole picture is visible at a glance. Choose the KPIs your first ninety days will actually move.
If a KPI has no baseline today, the first task in your plan is to set one.
17:00 – 22:00
The leadership model for AI survival
- Everything resolves, the book says, into a leadership problem: this industry's AI cannot be led by a pure technologist or a pure operator. Eight capabilities define the leader who can.
- Strategic clarity: knowing where AI creates value and refusing to chase it where it does not. Operational discipline: linking every model to a real decision and action in the field. Data accountability: making data ownership non-negotiable. Governance maturity: controlling risk, cyber exposure and model trust.
- People development: building the workforce. Commercial focus: anchoring every investment to a baseline and a business outcome, and killing what does not pay. Execution speed: moving from pilots to scale through disciplined replication. Local capability: an intelligence layer owned and operated by Nigerians, not rented.
- THE DOCTRINE
The decisive leader is not the most technical person in the room. It is the one who can turn technology into business performance by connecting all eight.
- IN THE FIELD · THE WORKED CASE
- Field. In a fictional fabrication yard, the book's priorities point to welding-productivity tracking and material traceability. The field question is whether the yard's records of welds, inspections and materials are good enough to track at all, so the first KPI it can honestly baseline may be data-quality score.
IN THE BOARDROOM · THE PARALLEL- Boardroom. The same yard's board chooses a focused KPI set across the four themes, each with a baseline and an owner. Commercial focus means it agrees now what result would make it stop the work.
Score yourself on the eight before your board does.
22:00 – 25:00
Before you close the laptop
- Write down your company type, its objective in the book's words, and the KPIs you will put in front of the board.
- This session produces Section 28 of your 90-Day AI Survival Game Plan: the game plan for your company type and the KPIs you will track.
- Pop Quiz 6 opens now: five questions on Sessions 26 to 28, closed book. Sit it before Session 29.
Pick fewer KPIs and baseline all of them, rather than more with blanks.
Key Terms
Outcome metric
A business result moved against a baseline, such as downtime reduction, as opposed to an activity metric such as dashboards built.
MTBF and MTTR
Mean time between failures and mean time to repair, which the book tracks underneath its reliability-and-operations KPIs rather than as separate lines.
Bid/no-bid discipline
Deciding before effort is spent which tenders are winnable and profitable.
Integrative leadership
Connecting operations, data, risk, technology, finance, people and governance into one executable operating model.
Self-Check — Before You Submit
1. The managing director of a fictional marine and logistics company wants the first AI investment to be a fleet-wide digital twin. Vessel downtime is the loss that costs the company most. What do you advise?
- A. Approve the twin, since twins suit marine operations
- B. Start with vessel-maintenance prediction tied to utilisation and downtime, with a baseline, as the book's plan for the type directs
- C. Start with demand forecasting
- D. Wait until all sixteen KPIs have baselines
B. Start with vessel-maintenance prediction tied to utilisation and downtime, with a baseline, as the book's plan for the type directs — The book's marine priorities lead with vessel-maintenance prediction, and its objective is to improve utilisation and reduce downtime. Focus means starting where value is clearest. Twins do matter for marine operations, the book says, but they are not on its marine priority list, and demand forecasting is a downstream priority.
2. Which KPI belongs to the book's capability theme?
- A. Tender win-rate improvement
- B. Data-quality score
- C. Equipment availability
- D. AI adoption rate
D. AI adoption rate — The capability theme holds AI adoption rate and workforce-training completion. Data-quality score is tempting because it sounds like capability, but the book places it under compliance and data.
3. Why does the book put project controls and schedule-risk prediction among the first priorities for EPC and construction companies?
- A. In project businesses the unforecast surprise is exactly where margin disappears
- B. NUPRC requires them of contractors
- C. They are the cheapest tools to buy
- D. They replace the project manager
A. In project businesses the unforecast surprise is exactly where margin disappears — The book's EPC objective is faster delivery with stronger cost control and fewer execution surprises. It cites no regulator requirement for these tools.
Today’s Assignment
Write Section 28 of your plan: the game plan for your company's type and the KPIs your board will track.
- Name your company type and state its objective as the book gives it. If your business spans two types, choose one and explain why, by reference to your costliest loss.
- From the book's priorities for your type, select the two or three you will act on first, and reconcile them with your three pillars from Section 27. Where they disagree, say which you are following and why.
- Choose a focused set of KPIs from the sixteen, across the four themes. For each, give today's baseline, or state that it is not yet known and who will establish it by day 30, with the owner by role and the source of the number.
- Name the activity metrics your company currently reports to its board in place of outcomes.
- Score yourself against the eight leadership capabilities and name the two weakest, with what you will do about each. If your company has no AI in use yet, the KPIs still apply: they measure the operation, not the software.
YOUR 90-DAY PLAN
Section 28 — The game plan for our company type and the KPIs we will track.
How this is marked — the rubric
| Company type and the book's objective stated, with a reasoned choice where types overlap | 15 |
| Two or three priorities chosen and reconciled with the Section 27 pillars | 25 |
| A focused KPI set across all four themes, each with baseline or honest gap, owner and source | 35 |
| Activity metrics named honestly and the two weakest leadership capabilities addressed | 25 |
| Total · pass mark 70 | 100 |
Submitting session 28
Write your answer in Word or Google Docs. The first line must read
Write your answer in Word or Google Docs. The first line must read
Session: 28, or the submission is refused. Upload the Word file (.docx), or switch to the Google Doc tab and paste your document’s share link.Marked against the rubric above and returned as a written letter, saved to your progress page. Do not submit until you have passed the self-check above.ASSESSMENT NOW OPEN
Pop Quiz 6 — The Game Plan opens now. Sit it before you start the next session — it is how you find out whether the doctrine landed, and the practice papers may be resat as often as you like.
Sit Pop Quiz 6 →Closed book, no notes, and Coach Ritchie switches himself off while you sit it.Capabilities Covered
Segment-specific sequencingKPI selectionBaseline disciplineLeadership self-assessmentOutcome over activity
Part Six · The Leadership MandateSESSION 29
The 90-Day AI Survival Roadmap
Why This Matters
Twenty-eight sections describe what your company should do. This one says when, and who. Chapter 17 compresses the start into a 90-day readiness sprint: thirty days to diagnose the company honestly, thirty to design the AI operating model, thirty to launch three to five pilots that each carry a measured business outcome. It is not the full transformation. It is the starting discipline that proves your company can convert intelligence into action, and it earns the mandate for everything after. Appendix A gives you the blueprint each pilot follows; Appendix B gives you the board metrics that judge the whole effort. Practice Exam 6 opens at the end.
25-Minute Module
0:00 – 5:00
A sprint, not the transformation
- The sprint is engineered to produce evidence fast, because momentum in transformation comes from visible, measured wins, not from a multi-year programme that shows nothing until the end.
Days Phase Deliverable 1–30 Diagnose the company, honestly An AI Readiness Diagnostic report 31–60 Design the AI operating model The AI operating model and an AI Program Office, owned by the business 61–90 Launch pilots that matter Three to five measured pilots live - Appendix A adds the shape of every pilot. The book says every domain is executed through the same disciplined ninety-day proof: discover and baseline, build in context, adopt and measure. Read the two together: the sprint is the company's ninety days, and Appendix A is the shape of each pilot inside it and beyond it.
- The blueprint, the book says, is deliberately simple, because its power is in being run, not admired.
If your first ninety days could pass without a single measured result, redesign them.
5:00 – 11:00
Days 1–30: diagnose the company, honestly
- Read the current state without flattering anyone: data sources and quality gaps, existing systems, field-reporting, maintenance and production processes, tendering and compliance workflows, cybersecurity maturity, workforce AI literacy, leadership readiness, and the real bottlenecks where value is leaking.
- The AI Readiness Diagnostic answers five questions: where is value being lost; what data exists today; what data cannot be trusted; which use cases create the fastest value; and what must be fixed before scaling.
- THE DOCTRINE
A diagnostic that protects egos is worthless. The one that names the real gaps is what makes the next sixty days productive.
- For the first pilot, Appendix A's opening month does the same work at small scale: select one high-value, bounded problem the organisation already complains about; map how the work actually flows today; quantify the cost of its friction; name the owners; and define the single decision the pilot will improve. Output: a written charter with scope, baseline, data required and an accountability map.
Ask someone who does not report to you to read the diagnostic before you sign it.
11:00 – 17:00
Days 31–60: design the AI operating model
- Structure: a governance committee, a data-ownership model, a prioritised use-case portfolio, the priority pilots, integrations, cybersecurity controls, a training plan, a vendor and partner strategy, a budget, success metrics, risk controls and an implementation timeline.
- Establish an AI Program Office, small but empowered, connecting IT, operations, finance, commercial, HSE, compliance, HR and procurement under executive sponsorship.
- THE DOCTRINE
AI cannot belong to IT alone. An AI programme owned by IT produces tools; an AI programme owned by the business produces performance.
- Appendix A's second month for a pilot: connect the necessary data, stand up the decision view, define the new workflow with its owners and controls, and review it with the people who will use it. The objective is a working capability usable against real conditions, not the full architecture.
- IN THE FIELD · THE WORKED CASE
- Field. A fictional downstream company's first pilot is inventory optimisation at its depots. In month two the depot team tests the decision view against how stock is actually received, finds two sites record volumes differently, and sends that gap back to the data-ownership model before going live.
IN THE BOARDROOM · THE PARALLEL- Boardroom. IT offers to run the programme; the executive committee places it with the commercial director, with IT, finance and HSE in the Program Office, because only the business can decide what is worth doing and be held accountable.
If the Program Office cannot name a business owner for every pilot, it is not yet designed.
17:00 – 22:00
Days 61–90: launch pilots that matter
- Restraint is the discipline. Do not start with twenty; start with three to five, chosen for clear value and adequate data. The book's candidates include predictive maintenance on critical equipment, AI-assisted tender compliance and bid readiness, field-reporting automation, production-deferment analysis, compliance-document tracking, procurement-delay risk alerting, pipeline-surveillance risk mapping and an executive operations dashboard.
- Each pilot needs nine things: a business owner, a data owner, a success metric, a timeline, a budget, a risk review, a governance control, an executive sponsor and a measured outcome. Appendix A's final month runs the workflow live, tracks decisions to outcomes, measures against the baseline and ends in a decision: continue, expand, adjust or stop.
- THE DOCTRINE
A pilot without a measured business outcome is not a pilot. It is a demonstration, and demonstrations do not compound into a company.
Appendix B board metric What it shows Decision cycle time The master metric: signal to decision to action Production reliability How dependably planned production is delivered Operating-margin improvement Improvement attributable to the intelligent operating model Workflow automation rate Routine work executed through governed, automated workflows Reporting accuracy The ability to produce, trust and defend operational, financial and regulatory numbers AI adoption quality Whether AI is used well, in real decisions, and trusted by its users Cross-functional accountability Whether ownership of outcomes that span silos is clear and honoured Strategy execution velocity The rate at which intention becomes accomplished change - Each metric must be tied to a baseline. A transformation that cannot state today's decision cycle time cannot credibly claim to have improved it.
Think of decision cycle time as your Operating-Intelligence Loop with a clock on it. Know yours before day 1.
22:00 – 25:00
Before you close the laptop
- Write down your three to five pilots and, for the first one, the single decision it will improve.
- This session produces Section 29 of your 90-Day AI Survival Game Plan: your ninety days, diagnose, design and launch, with owners and the stop-or-scale decision.
- Practice Exam 6 opens now: fifteen questions across the full syllabus, Sessions 1 to 29, closed book. Sit it before Session 30, where the Final Exam waits.
Put the stop-or-scale review in the diary today, with the names of the people who will be in the room.
Key Terms
AI Readiness Diagnostic
The day-30 deliverable: an honest read of the company answering where value is lost, what data exists, what cannot be trusted, which uses create value fastest and what must be fixed before scaling.
AI Program Office
A small, empowered unit connecting IT, operations, finance, commercial, HSE, compliance, HR and procurement under executive sponsorship, owned by the business rather than IT.
Decision cycle time
Appendix B's master metric: the elapsed time from signal to decision to action.
Demonstration
In the book's sense, a pilot without a measured business outcome. It may impress, but it does not compound.
Self-Check — Before You Submit
1. On day 65 at a fictional EPC company, the programme manager proposes launching eleven pilots to show momentum. Several have no baseline or business owner. What do you decide?
- A. Launch all eleven; momentum matters most
- B. Launch now and add baselines later
- C. Cut to three to five with clear value and adequate data, each with all nine elements
- D. Postpone every pilot to next year
C. Cut to three to five with clear value and adequate data, each with all nine elements — The book says start with three to five, not twenty, each with a business owner and a measured outcome. Each pilot also needs a data owner and a success metric; launched without baselines or owners, pilots are demonstrations, and demonstrations do not compound.
2. Why does the book insist the AI programme is owned by the business rather than IT alone?
- A. IT lacks the budget
- B. Only the business can decide what is worth doing and be held accountable for the outcome
- C. The regulator requires it
- D. IT cannot run pilots
B. Only the business can decide what is worth doing and be held accountable for the outcome — The book calls ownership the single most important design decision: IT-owned programmes produce tools, business-owned ones produce performance. It gives no budget or regulatory reason.
3. Which board metric does Appendix B call the master metric?
- A. AI adoption quality
- B. Workflow automation rate
- C. Operating-margin improvement
- D. Decision cycle time: signal to decision to action
D. Decision cycle time: signal to decision to action — Appendix B names decision cycle time. Operating-margin improvement is one of its other seven core indicators, not the master metric.
Today’s Assignment
Write Section 29 of your plan: your company's ninety days, diagnose, design and launch, with owners, baselines and the stop-or-scale decision.
- Days 1–30: who leads the AI Readiness Diagnostic, what it examines, and your honest draft answers today to its five questions. Name what in your company is most likely to make it flatter than the truth.
- Days 31–60: your governance committee and Program Office by role, the executive sponsor, the data-ownership model and the business function that owns the programme.
- Days 61–90: three to five pilots drawn from Sections 26 to 28, each with the nine elements.
- A charter for the first pilot in Appendix A form: scope, baseline, data required, accountability map and the single decision it improves, with today's decision cycle time for that decision or how you will measure it.
- The continue, expand, adjust or stop decision: its date, who makes it and what result triggers each outcome. If your company has no AI in use yet, the ninety days still start with the diagnostic.
YOUR 90-DAY PLAN
Section 29 — Our ninety days: diagnose, design, launch, with owners and the stop-or-scale decision.
How this is marked — the rubric
| An honest diagnostic plan with draft answers to all five questions | 20 |
| An operating model owned by the business, with committee, Program Office and sponsor by role | 20 |
| Three to five pilots, each with all nine elements | 25 |
| A first-pilot charter in Appendix A form, with baseline and decision cycle time | 20 |
| A dated stop-or-scale decision with criteria for each outcome | 15 |
| Total · pass mark 70 | 100 |
Submitting session 29
Write your answer in Word or Google Docs. The first line must read
Write your answer in Word or Google Docs. The first line must read
Session: 29, or the submission is refused. Upload the Word file (.docx), or switch to the Google Doc tab and paste your document’s share link.Marked against the rubric above and returned as a written letter, saved to your progress page. Do not submit until you have passed the self-check above.ASSESSMENT NOW OPEN
Practice Exam 6 — Full Syllabus opens now. Sit it before you start the next session — it is how you find out whether the doctrine landed, and the practice papers may be resat as often as you like.
Sit Practice Exam 6 →Closed book, no notes, and Coach Ritchie switches himself off while you sit it.Capabilities Covered
Readiness diagnosisAI operating model designPilot portfolio disciplineBoard-level metricsStop-or-scale decisions
CapstoneSESSION 30
Move Now: The Signed 90-Day AI Survival Game Plan
Why This Matters
Twenty-nine sessions ago you named your company, its type and the one loss that costs it most. Every section since has been written around that loss. Today you assemble them into one document, sign it and put it before your executive committee: your 90-Day AI Survival Game Plan. The book ends on two instructions, and this session carries both: Chapter 17's final message, move now, and the conclusion's charge, which belongs to you rather than to a vendor or a digital team. After you submit the plan, you sit the Final Exam.
25-Minute Module
0:00 – 5:00
The final message: move now
- AI is not waiting for Nigerian oil and gas companies to get comfortable. The regulatory environment is becoming measurement-based, margins are tighter, assets are older and production ambitions are climbing. Companies that wait will be forced to react on someone else's timetable; those that move now will set the standard.
- The path, in the book's words, is the whole series compressed into action: build executive AI fluency, fix the data foundation, create real operational visibility, prioritise predictive reliability, optimise production decisions, deploy disciplined twins on mature assets, sharpen tender and commercial intelligence, protect pipelines with predictive surveillance, make compliance and MRV audit-grade, govern AI rigorously, build Nigerian AI capability and assemble an implementation ecosystem.
- THE DOCTRINE
None of it is exotic; all of it is achievable; sequenced correctly, it is decisive.
The method in one line: data into intelligence, intelligence into decisions, decisions into action, action into measurable value.
5:00 – 10:00
Leadership using AI
- The argument of the book is one sentence: artificial intelligence will not save Nigerian oil and gas; leadership using artificial intelligence will.
- THE DOCTRINE
Every ingredient is present except the one that cannot be bought: the executive discipline to sequence it. Data first, bounded pilots, human judgement in the loop where the stakes are high, scale what pays.
- That discipline cannot be delegated to a vendor or to a chief digital officer without executive air cover. It belongs to leaders willing to be held to the number.
- I began the book with a moment of complacency: an offer to speak on AI at one of the industry's flagship gatherings, answered politely with the verdict that a session on AI was not needed. My answer is direct. It is needed, and urgently, because the window in which AI is a differentiator rather than table stakes is open now and closing.
- IN THE FIELD · THE WORKED CASE
- Field. At the asset, moving now means the one loss you named in Session 1 has an owner, a baseline and a start date.
IN THE BOARDROOM · THE PARALLEL- Boardroom. At the table, it means a signed plan, agreed decision dates and a signatory held to the number. The same loop, seen from both ends.
Knowing is not the same as leading. The signature is where one becomes the other.
10:00 – 17:00
The plan, section by section
- Assemble the twenty-nine sections in this order, each under its own heading.
- The inflection. 1. Our company, its type, the one loss that costs us most, and where our Operating-Intelligence Loop breaks. 2. The value at stake and our data-to-decision conversion gap. 3. Two candidate uses on the value chain, and the leadership priorities that apply. 4. How our operating model changes, function by function. 5. Our commercial value map, executive operating rhythm and board metrics with baselines.
- The foundations. 6. Data readiness for our first decision. 7. Our signal-to-decision path and the pilot outline that tests it. 8. What our first use case requires of the engine. 9. Build, buy or partner, observability and the funding case for our engine. 10. Where expert judgement must be accelerated and never replaced in our business.
- Intelligence in the field. 11. Our maturity stage and the credibility of the numbers we report. 12. Whether a living twin is warranted, what it must answer, and its pilot. 13. Our thin-margin discipline: the avoidable losses ranked by risk-adjusted impact. 14. Our binding constraint, the next one, and how AI would find them. 15. Where our reliability practice sits and the step up to predictive.
- Gas, the downstream and operational trust. 16. Our gas, flaring or emissions-measurement position and the AI response. 17. How the downstream reset changes our business. 18. Our accountability map across the three lines of defence. 19. Our governance pilot, rhythm and scoreboard. 20. Our largest preventable loss and the fused-signal response.
- Controlled autonomy, capital and people. 21. Three uses placed on the autonomy ladder, with guardrails. 22. Our first bounded agentic pilot and the board's questions answered. 23. Our credibility flywheel and what investors or lenders can verify today. 24. The numbers we must defend, their owners, and our capital-credibility pilot. 25. The capability each role in our company needs.
- The leadership mandate. 26. Our capability pilot and scoreboard. 27. Our company scored on the twelve pillars, and the three that come first. 28. The game plan for our company type and the KPIs we will track. 29. Our ninety days: diagnose, design, launch, with owners and the stop-or-scale decision.
Where a later section changed your mind about an earlier one, correct the earlier one and say so in a line. That is rigour, not inconsistency.
17:00 – 22:00
The executive summary and the signature
- On top goes a one-page executive summary: the company and its type; the one loss and its baseline; the three pillars that come first; the pilots, each with a business owner and measured outcome; KPIs and board metrics with baselines; the capability pilot; and the dates of the day-30, day-60 and day-90 deliverables and the stop-or-scale decision.
- Test the plan against the six frameworks. Does it close your loop where it breaks? Is the data foundation built for the first decision? Is every use placed on the Autonomy Ladder and the Value-versus-Risk Map? Is the order sequenced? Does every pilot connect the capability to the barrel?
- THE DOCTRINE
Pick the one loss that costs you most, name the owner, set the baseline, and start the ninety days on Monday. Not a committee. Not a strategy offsite. A start.
- Sign and date it. A signature on this plan means the people who read it may hold you to its numbers.
Sign only what you would defend to your own board on day 90.
22:00 – 25:00
The capstone and the Final Exam
- The capstone is the assembled, signed plan. It will run well beyond 900 words; the summary stays on one page. Send it even if some sections embarrass you. I would rather mark an honest document than a tidy one.
- Then sit the Final Exam on this page: fifteen questions, three rebuilt from questions you got wrong on the practice papers, closed book, pass mark 70.
- The technology will keep improving whether or not you act; the only variable you control is whether your company is built to use it. The train is leaving. Board it, and put Nigerian hands on the controls.
The work is hard. It is also worth it. Let us build it well, and build it ours.
Key Terms
90-Day AI Survival Game Plan
Sections 1 to 29 assembled and signed under a one-page executive summary, with owners and baselines.
Readiness sprint
The book's first ninety days: not the full transformation, but the discipline that proves the company can turn intelligence into action.
Executive air cover
Backing from the executive team. The book says the discipline cannot be delegated to a chief digital officer working without it.
Differentiator versus table stakes
AI that sets a company apart versus AI that is merely the price of competing. The book says the window in which AI is a differentiator is open now and closing.
Self-Check — Before You Submit
1. The CEO of a fictional marginal field operator reads the signed plan and proposes a strategy offsite next quarter before anything starts. What does the book's charge direct?
- A. Hold the offsite; alignment comes first
- B. Form a committee to review the plan
- C. Wait for the next budget cycle
- D. Name the loss owner, set the baseline and start the ninety days on Monday
D. Name the loss owner, set the baseline and start the ninety days on Monday — Chapter 17's charge is explicit: not a committee, not a strategy offsite, a start. The offsite feels responsible, but it is exactly the delay the book warns against.
2. What does the conclusion name as the one ingredient that cannot be bought?
- A. The executive discipline to sequence it
- B. Frontier AI models
- C. Cloud computing capacity
- D. Data scientists
A. The executive discipline to sequence it — The book says the technology is ready, the data gettable and the talent reachable. What is missing is the discipline to sequence it, which is why it is a leadership deliverable, not a purchase.
3. According to the book's final message, what happens to companies that wait?
- A. They avoid early mistakes
- B. They gain from cheaper technology later
- C. They are forced to react on someone else's timetable
- D. They are protected by the regulator
C. They are forced to react on someone else's timetable — The book says waiting companies will react on someone else's timetable while movers set the standard. Cheaper tools later is tempting, but the advantage compounds in decision quality, not tool prices.
Today’s Assignment
Assemble Sections 1 to 29 into one signed 90-Day AI Survival Game Plan for your company, with a one-page executive summary for your executive committee.
- Assemble all twenty-nine sections in the order listed in this session and edit for consistency, noting in one line wherever a later session changed an earlier section.
- Write the one-page executive summary with every element listed in this session.
- Test the plan against the six frameworks and record, in a short paragraph, where it is weakest.
- Name the executive sponsor and the owner of the one loss by role, and the Monday on which your company's ninety days begin.
- Sign and date the plan, and name the commitment you are least confident your company will keep, and why. The plan will run beyond 900 words; the summary may not. If your company has no AI in use yet, the plan still stands: it starts from the loss, not the software.
YOUR 90-DAY PLAN
The assembled, signed 90-Day AI Survival Game Plan.
How this is marked — the rubric
| All twenty-nine sections present, in order and consistent, with corrections noted | 25 |
| A one-page executive summary covering every required element | 25 |
| An honest framework test naming the weakest point | 15 |
| Sponsor and loss owner by role, a start date and dated 30, 60 and 90-day deliverables | 20 |
| Signed, with the least-certain commitment named and explained | 15 |
| Total · pass mark 70 | 100 |
Submitting session 30
Write your answer in Word or Google Docs. The first line must read
Write your answer in Word or Google Docs. The first line must read
Session: 30, or the submission is refused. Upload the Word file (.docx), or switch to the Google Doc tab and paste your document’s share link.Marked against the rubric above and returned as a written letter, saved to your progress page. Do not submit until you have passed the self-check above.FINAL EXAM
Your Final Exam is sat in this session. 15 questions, of which 3 are rebuilt from questions you got wrong on the practice papers. Sit it once your capstone is submitted.
Sit the Final Exam →Closed book, no notes, and Coach Ritchie switches himself off while you sit it.Capabilities Covered
Plan assemblyExecutive summary writingFramework self-testAccountable sponsorshipMoving now