THE INTELLIGENT PETROCHEMICAL COMPANY LEADERSHIP PROGRAMME
The decision behind the chemical.
20-Session Programme · 25 Minutes Per Day · Free · For leaders in Nigeria’s petrochemical and specialty chemical sector
WEEK 1 OF 4 · LEADERSHIP, THE BUSINESS MODEL AND FEEDSTOCK · SESSIONS 1–5 · WORK AT YOUR OWN PACE
Orientation · The Decision Behind the ChemicalSESSION 1
The Decision Behind the Chemical and the Reader's Toolkit
Why This Matters
A petrochemical company has to connect technical performance with commercial discipline. Feedstock, production, formulation, product quality and customer application each affect the result, and artificial intelligence earns its place only when it supports a decision within that pattern. Over twenty sessions you will build a signed 90-Day AI Leadership Plan for your own company around one recurring decision with a meaningful consequence, which you choose today. The question I will keep asking is the one the book asks of every proposal: which decision will change, who can authorise that change and how its effect will be measured. A recommendation that remains on a screen has not closed the operating loop. Closing it is leadership work, and that is what this programme trains.
25-Minute Module
0:00 – 4:00
The argument, and the finish line
- The starting point: an improvement in a technical measure does not automatically become an improvement in earnings, and a fluent explanation cannot replace validated information.
- THE DOCTRINE
Artificial intelligence becomes useful when it supports a decision. The executive still needs to know which decision will change, who can authorise that change and how its effect will be measured.
- The four weeks: 1 Leadership, the Business Model and Feedstock; 2 Production, Quality and the Customer; 3 Reliability, Resources and Enterprise Execution; 4 Authority, Governance and the First Ninety Days. The finish line: nineteen plan sections written one per session, assembled and signed in Session 20. House rules: write about your own company; bring a real decision; no value claim without a baseline; no decision without a named owner; report a blocker the same day. The pass mark is 70.
- Said once: this is leadership training, not investment, legal, regulatory or engineering advice. The six frameworks are the author's synthesis, not official standards. A recommendation involving formulation, product release or a customer treatment programme requires the relevant technical review.
Ask which decision will change before you ask what to buy.
4:00 – 8:00
Begin with the business you actually run
- The book distinguishes the continuous producer, the batch formulator, the specialist supplier and the supplier with technical service. A producer may study how feed composition affects conversion and the saleable mix; a formulator, how raw material variability connects to a recipe version; a supplier, whether an imported lot meets the specification. These are different decisions.
- Define where your company has evidence, where it has authority and where it depends on another party. Where a session covers an activity your business does not run, apply it to your nearest equivalent decision or to the counterpart you depend on. Nigeria is the setting, not a justification for sweeping assumptions: establish your own exposure to imported inputs, power interruptions, foreign currency and customer payment behaviour.
- Know what kind of intelligence is proposed. Blocking an unapproved specification is valuable without machine learning. The book reserves the term AI for systems that learn patterns, produce probabilistic predictions or generate content. NIST AI RMF 1.0 organises risk management around Govern, Map, Measure and Manage; the book treats it as an external reference for governance, while the six frameworks supply the operating vocabulary.
A reconciled stock ledger should not depend on the creativity of a language model.
8:00 – 17:00
The six frameworks, as questions
Framework In brief The question it forces 1. Operating-Intelligence Loop Sense, decide, act, learn Where does the loop break for our decision? 2. Data-Foundation Precondition Evidence adequate for the consequence Do we know the property, unit, method and specification behind a result? 3. Autonomy Ladder Inform, draft, recommend, execute approved steps What may this application read, create, change and send? 4. Value-versus-Risk Map Benefit and consequence of error assessed together Which failure mode needs a boundary regardless of return? 5. Sequencing Discipline Foundation, bounded pilot, oversight, expansion What must be demonstrated before the next step? 6. Operations-Excellence Lens Measures of the actual business Which measure have finance and operations agreed? - The loop stays open if the software merely issues an alert, and equally if management approves an investigation but supplies no sampling resources. The fix may be a faster approval or a funded action, not a more elaborate prediction.
- IN THE PLANT · THE WORKED CASE
- Plant. The book's illustrative quality exception: a batch's viscosity is outside the agreed specification. A system finds similar records; the laboratory confirms the result, production examines the batch history and the designated authority decides the disposition.
IN THE MARGIN · THE PARALLEL- Margin. Finance records the actual consequence, and the procedure is revised if the evidence supports it. If that consequence is never recorded, the loop breaks at learn. It is the same loop, seen from the money.
- The data floor and the ladder. A result marked 98 means little until the team knows the property, unit, test method, sample, product and specification. Keep missing information visible. Permission to draft a supplier query does not authorise releasing a lot or changing a formulation.
- THE DOCTRINE
Permission should be scaled to the organisation's ability to evaluate the action, not to the software's ability to produce it.
- The map, the sequence and the lens. Record failure modes separately, such as an unsuitable product released or a valid product unnecessarily held. A pilot needs an actual decision, an owner and an acceptance test. Lower energy per tonne of gross production may conceal more off-specification product.
Find the broken stage before you fund another analytical feature.
17:00 – 22:00
Carry one decision record, and make it testable
- Use six fields: decision requested, alternatives, evidence and uncertainty, accountable owner, cash and operating consequences, and next review date. The book's illustrative entry asks whether to introduce advisory lot-readiness checks for one product family; the owner is the quality manager and the pilot makes no release decisions.
- THE DOCTRINE
An executive framework earns its place when it produces a proposition that can fail.
- Write it in five sentences: the decision and last useful intervention time; the evidence available then; the feasible response and its approver; the comparison with the current arrangement; the outcome that justifies the full cost. Add a counterexample to every pilot charter.
- What an experienced reader should demand: the original decision paper, the recommendation, the action record and the financial reconciliation at the same review.
If you cannot identify the required record, that absence is already a finding.
22:00 – 25:00
Write it down, and Section 1
- Before you close the laptop, name your company's business model and the decision you will carry, with its alternatives, including retaining the current arrangement, the date a choice is needed and the person authorised to make it.
- Choose one recurring decision with a meaningful consequence and an achievable evaluation route: a supplier lot, a production campaign, an overdue customer trial or a recurring quality exception. The right answer may later prove to be a simpler control, or no model at all.
- Section 1 of your plan is Our company, its business model, the one decision we will carry and where its loop breaks.
The decision you choose today is the one you will sign for in Session 20.
Key Terms
Operating-Intelligence Loop
Sense, decide, act and learn; closed only when the outcome is examined against the evidence.
Data contract
Agreed fields, meanings, units, quality, timing, owners and treatment of missing information for a decision.
Autonomy
Permission for a system to carry out defined actions, separate from its analytical capability.
Decision record
Decision requested, alternatives, evidence and uncertainty, accountable owner, cash and operating consequences, and next review date.
Self-Check — Before You Submit
1. A fictional specialist supplier's application flags lots likely to miss release. Management approved investigations but supplied no sampling resources. Where has the loop broken?
- A. At sense: the model needs more data
- B. At decide: alternatives were never compared
- C. At act: nobody has the resources to respond
- D. At learn: the outcome was never reviewed
C. At act: nobody has the resources to respond — An approved investigation with no sampling resources leaves the loop open at act. Improving the model is tempting but does nothing for the failing stage.
2. Which does the book treat as valuable ordinary digital control rather than AI?
- A. Preventing an unapproved specification from being used
- B. A probabilistic forecast of supplier delay
- C. A generated draft explaining a quality exception
- D. Learning abnormal operating patterns from historian data
A. Preventing an unapproved specification from being used — Blocking an unapproved specification is valuable without machine learning. The other three learn patterns, predict probabilistically or generate content, which is how the book uses the term AI.
3. A proposal has a large benefit estimate, but one failure mode is an unsuitable product released. How does the Value-versus-Risk Map treat it?
- A. Average it into one score with the benefit
- B. Record it separately; some consequences need a protection or authority boundary regardless of return
- C. Accept it, because the benefit is large
- D. Leave it to the technology team
B. Record it separately; some consequences need a protection or authority boundary regardless of return — A single score can hide serious consequences, and a large benefit estimate does not make a proposal acceptable. Averaging feels balanced but conceals the release risk.
Today’s Assignment
Name your company and its business model, choose the recurring decision you will carry, trace it through the Operating-Intelligence Loop and state where the loop breaks.
- Your company and its model. What it sells, which of the four models it runs, and where it has evidence, authority and dependence on another party.
- The decision. One recurring decision with a meaningful consequence, its alternatives (including the current arrangement), the date a choice is needed and the authorised owner by role.
- The loop. Sense, decide, act and learn as they work today; the stage where it breaks, why, and who owns that stage.
- The honest part. What your business gets wrong about this decision today. No AI in use yet is normal; work from the decision in front of you. Four hundred to nine hundred words.
YOUR 90-DAY PLAN
Section 1 — Our company, its business model, the one decision we will carry and where its loop breaks.
How this is marked — the rubric
| Names the company, its business model and where it holds evidence, authority and dependence | 20 |
| States one recurring decision with alternatives, a decision date and an owner by role, not a department | 30 |
| Locates the loop break at a named stage with a reason and an owner | 25 |
| Distinguishes ordinary control from AI for the gap and names a real current weakness | 25 |
| Total · pass mark 70 | 100 |
Submitting session 1
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: 1, 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
Decision-first leadershipThe six toolkit frameworksLocating the broken loop stageWriting a decision recordFraming a testable proposition
Part One · Leadership and the Business ModelSESSION 2
Leading the Intelligent Petrochemical Company
Why This Matters
Picture three reports that do not agree: output above plan, finished goods growing, complaints and collections slipping. The book opens its leadership chapter with exactly that illustrative situation and calls it a leadership problem before it is an AI opportunity. In Session 1 you chose a decision. Today you decide what question it answers, who executes it, how your transformation function is directed and how your operating review will resolve the choice rather than revisit it. Without that, the most accurate model your company buys can produce alerts that belong to nobody. This is the part of the programme that only the sponsoring executive can do.
25-Minute Module
0:00 – 5:00
Own the operating question
- In the book's illustration, the additional output may be waiting for release, assigned to the wrong customers or tied to invoices that will remain unpaid. Each function has a plausible account; nobody can yet say whether the business has improved.
- State the question the programme will improve. Increasing profitable accepted deliveries is a different objective from maximising plant throughput. Reducing release delays is different from reducing the number of laboratory tests. Lowering inventory exposure is different from cutting stock regardless of service consequences.
- The distinction changes the data and authority required. A throughput model can remain within production analysis. A delivery decision needs laboratory readiness, warehouse status, transport availability and customer acceptance. A working capital decision adds credit terms, supplier commitments and the financing calendar.
- THE DOCTRINE
Leadership owns the cross-functional choice even when specialists own each contributing record.
- Write the objective so it can be challenged: measure, boundary, period and constraints. The sponsor must explain what will be funded, which decisions remain under existing technical authority and what outcome would make the work worth continuing.
If every department can define success differently, you have not yet set an objective.
5:00 – 10:00
Give the decision an executable owner
- A committee can review a proposal, but an action needs a person. If a model identifies a probable material shortage, someone must verify it, compare procurement options and obtain approval before the shortage occurs. Responsibility for the alert alone is insufficient.
Role What the book says it owns Quality The relevant release determination Procurement Supplier communication and purchasing execution within its mandate Treasury Confirmation of funding availability Operations The feasible production response Managing director Tradeoffs that exceed delegated authority - An escalation path identifies the decision waiting, its deadline and the consequence of delay. It should not consist solely of copying more people into an email.
- If a purchasing recommendation requires a funding choice, the treasury decision belongs on the management agenda, with alternatives and cash timing.
Separate the person who decides from the people who supply the evidence.
10:00 – 15:00
Direct the transformation function, and make the review resolve
- Hiring a digital leader does not transfer operating accountability. The digital function can design information flows, evaluate systems and coordinate development. It cannot establish a product's release authority or supply money the company has not approved.
- Give it a specific mandate: make selected decisions better supported and more executable, working with the outcome owner, the specialist who can evaluate the evidence and finance. Its delivery should include a usable workflow and support arrangement, not simply a completed model.
- THE DOCTRINE
Operating leaders must provide access to records and time for evaluation. A sponsor who demands results while refusing either obligation has approved an expectation rather than a programme.
- Begin the operating review with changes requiring a decision; for each, require the decision requested, alternatives and evidence, and end with an owner, a resource commitment and a review date. Record a decision to defer and the evidence required to reopen it.
- Examine failed recommendations. If staff consistently ignore an alert, determine whether it is unreliable, too late, duplicative or impossible to act on. Adoption statistics alone cannot answer that.
A meeting that ends without an owner and a date has only shared information.
15:00 – 21:00
Measure management's contribution, and keep two reviews apart
- Distinguish system performance from organisational response. A useful prediction can fail commercially because an approved part was not purchased. Track the interval from evidence to decision and from approval to action, and examine overdue decisions by cause. Count verified operating benefit, cost of support, unresolved high-consequence exceptions and exposure to unsupported permissions. A product price increase or an unrelated maintenance intervention may explain a gain that looks like an AI result.
- IN THE PLANT · THE WORKED CASE
- Plant. The book's illustrative bridge: an earlier quality warning enables inspection before packaging. The technical review must show the warning is sufficiently reliable; the operating review must show staff can respond before packaging begins.
IN THE MARGIN · THE PARALLEL- Margin. Finance compares avoided packaging or rework expenditure against extra investigations and support cost. Crediting the programme with an entire batch's selling price would obscure the bridge.
- THE DOCTRINE
An intervention can be technically acceptable and economically unattractive, or financially attractive on paper and operationally infeasible.
- The book's illustrative intervention: a chemical supplier cannot say whether an urgent order can be delivered. Sales sees stock on a spreadsheet, the warehouse sees drums awaiting receipt inspection, quality awaits documentation and finance has placed the customer on credit hold. Management first commissions a shared order-readiness record using ordinary workflow controls, then tests an advisory application that cannot release stock, remove the credit hold or revise a delivery promise without approval.
I would reject an investment paper that cannot explain the decision the system improves.
21:00 – 25:00
Write it down, and Section 2
- Before you close the laptop, rewrite your Session 1 decision as an operating objective with a measure, boundary, period and constraints, and name its executable owner by role.
- Make the objective identify the tradeoff, as the book's examples do: more fulfilled orders within an approved cash limit, or more contribution from a constrained campaign window while maintaining quality and process protections.
- Section 2 of your plan is The operating question, its executable owner and the review that resolves it.
The question you own today is the one the board will hold you to.
Key Terms
Operating question
The specific objective the programme will improve, stated with measure, boundary, period and constraints so that it can be challenged.
Executable owner
The person who can verify information, compare options and obtain approval in time to act, as distinct from those supplying evidence.
Escalation path
A route naming the decision waiting, its deadline and the consequence of delay.
Operating control and investment accountability
Two separate reviews: whether a response is technically defensible within its scope, and whether quantities, costs and cash support the investment.
Self-Check — Before You Submit
1. Staff at a fictional batch formulator have ignored an early-warning alert for weeks. The digital lead proposes retraining the model. What should the operating review establish first?
- A. The adoption rate across the plant
- B. Whether the vendor offers a newer model
- C. Whether to make use of the alert mandatory
- D. Whether the alert is unreliable, too late, duplicative or impossible to act on
D. Whether the alert is unreliable, too late, duplicative or impossible to act on — The book says an ignored alert must be examined for those four causes, and that adoption statistics alone cannot answer the question. Retraining or mandating use assumes the cause before the action log has been read.
2. A purchasing recommendation requires a funding choice. Where does the book place that decision?
- A. With procurement, inside its purchasing mandate
- B. On the management agenda as a treasury decision, with alternatives and cash timing
- C. With the digital function that produced the recommendation
- D. In an email copied to every affected manager
B. On the management agenda as a treasury decision, with alternatives and cash timing — Procurement executes within its mandate, but a funding choice belongs on the management agenda with alternatives and cash timing. An escalation path should not consist solely of copying more people into an email.
3. An earlier quality warning lets a fictional plant inspect a batch before packaging. What benefit should finance credit?
- A. The entire batch's selling price
- B. The model's accuracy multiplied by monthly output
- C. Avoided packaging or rework expenditure, set against extra investigations and support cost
- D. Nothing until the board approves the programme
C. Avoided packaging or rework expenditure, set against extra investigations and support cost — The book's bridge compares avoided packaging or rework expenditure with extra investigations and support cost. Crediting the whole batch's selling price is the tempting error; it obscures that bridge.
Today’s Assignment
For your decision, own the operating question, name its executable owner, set how your transformation function is directed and how your operating review will resolve the choice, and separate operating control from investment accountability.
- The operating question. Measure, boundary, period, constraints and the tradeoff it accepts; what will be funded and which decisions stay under existing technical authority.
- The executable owner. The accountable owner by role, the evidence providers, and the escalation path with deadline and consequence of delay.
- The transformation function. Its mandate for this decision, what your operating leaders will supply in records and time, and what it must deliver beyond a model.
- The review. How your operating review will present this decision, what it must end with, and how deferrals and ignored recommendations are recorded.
- Two reviews. What the technical reviewer and finance must each establish, and one external movement that could be mistaken for an AI result in your business.
- The honest part. Where decisions in your company currently wait, and why. Four hundred to nine hundred words.
YOUR 90-DAY PLAN
Section 2 — The operating question, its executable owner and the review that resolves it.
How this is marked — the rubric
| States the objective with measure, boundary, period, constraints and its tradeoff | 25 |
| Names an accountable owner by role, the evidence providers and a specific escalation path | 25 |
| Sets a review that ends with an owner, resource commitment and date, and records deferrals and ignored alerts | 25 |
| Separates technical from financial evidence and names a real current delay in the company | 25 |
| Total · pass mark 70 | 100 |
Submitting session 2
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: 2, 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
Owning the operating questionAssigning executable ownershipDirecting the digital functionRunning a review that resolvesSeparating technical and financial evidence
Part One · Leadership and the Business ModelSESSION 3
Production, Formulation and Supply Business Models
Why This Matters
The phrase chemical company can conceal several different businesses. Some convert feedstock through a manufacturing process; some formulate to a controlled specification; some import and distribute; some combine supply with technical service at the customer's site. Each needs its own decision boundary, and an AI portfolio assembled without that distinction can become expensive and irrelevant. Having named the operating question and its owner in Session 2, today you write down what kind of business your company actually is: where it creates value, where it bears technical responsibility and where it commits cash. That record will let you reject unsuitable applications early, before your company buys a platform it cannot use.
25-Minute Module
0:00 – 7:00
Four business models, four decision boundaries
Model Where the decision lives Boundary to preserve Continuous producer A connected physical system: feed composition, process conditions, equipment, utilities, downstream capacity Analysis kept apart from process control; engineering review and change control govern any operating change Batch formulator Approved recipe, actual lots, quantities, sequence, equipment, conditions, laboratory result The formula is intellectual property; a raw material substitution enters validation and approval Specialist supplier Suitable product, identity and condition preserved, reliable delivery Manufacturer's certificate, own incoming checks and customer application test answer different questions Supplier with technical service Application, operating conditions, acceptance criteria, customer authority Whether you sell a product, an application service or a performance commitment - The producer. A decision that improves one unit can transfer a constraint elsewhere. A laboratory result may describe material produced before the sample reached the laboratory. A model used to examine options should not acquire permission to write to a control system by convenience.
- The formulator. Two batches bearing the same product name may have different raw material histories, and an average of their outcomes can conceal why one failed. Access should follow the purpose of the application.
- The supplier, with or without service. A product that conforms to a stated specification is not automatically proven for a particular application. If information is missing, an assistant should request it rather than invent a fit. The model cannot create evidence that the operating relationship does not supply.
- THE DOCTRINE
Leadership should define where the company creates value, where it bears technical responsibility and where it commits cash.
Name your business model before you name your application.
7:00 – 11:00
Connect the group without flattening the differences
- A group may centralise purchasing, treasury and digital support, and shared product and supplier identities are useful. Technical decisions still need the evidence and authority relevant to each business.
- Do not judge a producer, a formulator and a supplier by a single inventory days target without examining their constraints. A critical feedstock, a customer-specific additive and a slow-moving stock item may need different replenishment rules. Manufacturing conformity, dispatch conformity and customer application acceptance are different events; a group scorecard should preserve them and show how they connect to collected revenue.
- IN THE PLANT · THE WORKED CASE
- Plant. The book's illustrative portfolio review: a group's blending site has frequent rework on one product family. That decision needs reliable recipe versions, material lot links and test history.
IN THE MARGIN · THE PARALLEL- Margin. The group's importing business has excess stock in low-demand items and shortages in committed customer products. It needs committed demand, usable stock status, supplier lead-time records and credit terms. A common platform may offer an economic advantage if it reduces support cost and preserves the required boundaries.
Let the platform follow the operating design, not the other way round.
11:00 – 17:00
Locate the business in the value chain, and define yield
- The book's operating map separates conversion into basic chemical intermediates, production of polymers or resins, formulation of purchased ingredients, distribution of qualified products and technical service at the customer's application. These are analytical categories for leadership; they do not determine regulatory classification.
Measure The question it answers Reaction conversion The extent to which a reactant has reacted under the stated basis Selectivity The share directed toward a chosen product under a defined chemical basis Recovery How much of a relevant material is recovered through a stated boundary Accepted commercial yield Output that meets its approved disposition and commercial conditions - THE DOCTRINE
These yield measures answer different questions and should not be substituted for one another in an investment paper.
- Three illustrative traps. A model predicts more of a desired intermediate, but downstream separation cannot recover it at specification. A formulator's batch mass rises through additional carrier, which says nothing about customer performance or contribution. A distributor's stock rises through a larger import, but availability has improved only for products and dates the customer can accept.
Ask which yield before you ask how much.
17:00 – 21:00
Choose a method that fits the decision
- Deterministic checks for exact identity, approved status and arithmetic; forecasting for quantities or timings that must be estimated; optimisation for choosing among feasible alternatives under constraints; generative retrieval for locating and explaining controlled documents. They may work together, but evaluate them separately.
- A production schedule can be optimised with mathematical programming without a language model. A soft sensor estimates a property not directly available at the required decision time. An agent adds the ability to invoke actions, which creates an authority question beyond the quality of the explanation.
- Lawrence and colleagues' review of machine learning for industrial sensing and control (Control Engineering Practice, 2024) treats soft sensing, hybrid modelling, optimisation and control as related areas and emphasises domain knowledge alongside learned relationships; the book's source check reviewed the relevant sections of the authors' preprint. The book notes it does not establish a return for a Nigerian facility or a general permission for autonomous control.
Evaluate each component on its own terms, even when they share one screen.
21:00 – 25:00
Write it down, Section 3, and Pop Quiz 1
- Draft your one-page business-model record: product or service sold, technical responsibility retained, evidence available, point of customer acceptance, cash commitment and main constraint, plus the decisions your business can authorise and those requiring another party's approval.
- Add the transformation performed, the material boundary, the constrained resource, the analytical capability proposed and the information it will use.
- Section 3 of your plan is Our business-model record, our place in the value chain and the method that fits.
- Pop Quiz 1 opens now: five questions, closed book, covering Sessions 1 to 3. Sit it before Session 4.
A model should address a decision your company can actually influence.
Key Terms
Business-model record
A one-page record of what a business sells, the technical responsibility it retains, its evidence, acceptance point, cash commitment, main constraint and decision authority.
Accepted product
Product that has met the applicable authorised disposition and customer or commercial conditions within the stated measure.
Soft sensor
An estimate of a property from other available measurements or inputs, with its approved decision purpose made explicit.
Hybrid model
A model combining domain or physical relationships with a data-driven component, evaluated as a combined arrangement within its scope.
Self-Check — Before You Submit
1. A fictional group with a continuous plant, a blending site and an import business proposes one inventory days target for all three. What is the right response?
- A. Adopt it, because one target simplifies the board report
- B. Examine each business's constraints; different materials may need different replenishment rules
- C. Set the target at the importing business's level
- D. Replace the target with a single AI platform
B. Examine each business's constraints; different materials may need different replenishment rules — The book warns against judging different models by a single inventory days target without examining their constraints. Simpler reporting is the tempting argument, but central reporting should expose the reason for each commitment.
2. Which measure belongs in an investment paper about output that meets its approved disposition and commercial conditions?
- A. Reaction conversion
- B. Selectivity
- C. Recovery
- D. Accepted commercial yield
D. Accepted commercial yield — Accepted commercial yield concerns output meeting its approved disposition and commercial conditions. Conversion, selectivity and recovery answer different technical questions and should not be substituted for it.
3. A customer asks a fictional specialist supplier whether a product conforming to its stated specification suits the customer's application. What does the book say?
- A. Conformity does not prove suitability; the certificate, incoming checks and application test answer different questions
- B. Conformity proves suitability for any stated use
- C. The manufacturer's certificate settles it
- D. An assistant can infer suitability from similar products
A. Conformity does not prove suitability; the certificate, incoming checks and application test answer different questions — A conforming product is not automatically proven for a particular application. Relying on the certificate is the tempting shortcut; an assistant should request missing evidence rather than invent a fit.
Today’s Assignment
Write the business-model record for your company, locate it in the chemical value chain and state which method fits the decision you are carrying.
- The record. Every field above, including the decisions your business can authorise and those needing another party's approval.
- The value chain. Where your company sits on the book's operating map, and which yield or acceptance measure applies to your decision. If you are a group, say how the businesses differ and what they should share.
- The method. Deterministic check, forecast, optimisation, generative retrieval or a combination, and how each part would be evaluated separately.
- A mismatch. One application your company has considered, or could be offered, that this record would reject, and why.
- The honest part. Where your business currently blurs its model, its yield measure or its authority. Four hundred to nine hundred words.
YOUR 90-DAY PLAN
Section 3 — Our business-model record, our place in the value chain and the method that fits.
How this is marked — the rubric
| Completes every field of the business-model record, including decisions needing another party's approval | 30 |
| Locates the company on the operating map and names the correct yield or acceptance measure | 25 |
| Matches the method to the decision and keeps each component's evaluation separate | 25 |
| Rejects one unsuitable application with a reason drawn from the record | 20 |
| Total · pass mark 70 | 100 |
Submitting session 3
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: 3, 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 1 — The Operating Question and the Business Model 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 1 →Closed book, no notes, and Coach Ritchie switches himself off while you sit it.Capabilities Covered
Distinguishing business modelsWriting a business-model recordReading yield correctlyMatching method to decisionGroup design without flattening
Part One · Leadership and the Business ModelSESSION 4
Decisions That Connect Technical Performance to Margin
Why This Matters
A chemical business earns through accepted products and services delivered on terms it can fund. A proposal can report a technical gain without showing what that gain becomes in contribution and cash. This session gives you the questions that expose the difference: the correct denominator, the alternative that would have happened anyway, the bridge from tonnes to naira and the funding calendar behind it. Having written your business-model record in Session 3, you now connect your decision to money. If your finance and technical leads cannot reconstruct the bridge together, the case is not ready, however good the demonstration looked.
25-Minute Module
0:00 – 5:00
Ask for the bridge, and use the correct denominator
- Ask for a bridge from the physical change to the financial result. If yield improved, what additional quantity was accepted and sold? If rework fell, which cost was avoided? If release was faster, did invoicing change and was cash collected sooner?
- THE DOCTRINE
A chemical business earns through accepted products and services delivered on terms it can fund. A technical improvement matters commercially when it changes a quantity, cost, price, timing or obligation within that boundary.
- Gross output divided by input is not necessarily saleable product divided by input. A formulator may monitor first-pass conformity; a supplier may monitor accepted deliveries. Additional low-value by-product may increase physical yield while reducing contribution.
- Separate an accounting allocation from an incremental cost. A campaign may look unattractive after allocated fixed overhead yet still make a positive contribution to available capacity; a positive contribution does not cover the full long-term cost of running the business. State which question the calculation answers.
Before you compare two numbers, ask what each was divided by.
5:00 – 12:00
The assumed campaign calculation
- Plain language first. Contribution is revenue less the costs included in an expressly stated boundary. It is not full accounting profit.
- The book's teaching calculation is assumed and is not a market benchmark: a 100-tonne input campaign, material cost ₦70 million, standard product at ₦1 million per tonne, permitted lower grade at ₦600,000 per tonne, conversion ₦15 million, logistics ₦2 million. All permitted product is sold and collected within the period, with no tax, financing or fixed-overhead allocation.
Line (assumed) Baseline Technically approved intervention Standard product 92 t × ₦1 million = ₦92 million 94 t × ₦1 million = ₦94 million Permitted lower grade 5 t × ₦600,000 = ₦3 million 4 t × ₦600,000 = ₦2.4 million Loss 3 t 2 t Revenue ₦95 million ₦96.4 million Material, conversion, logistics ₦70m + ₦15m + ₦2m = ₦87 million ₦70m + ₦15.8m + ₦2m = ₦87.8 million Contribution ₦8 million ₦8.6 million - The incremental contribution is ₦600,000, not the ₦2 million value of two additional standard tonnes. Line by line: two more standard tonnes add ₦2 million, one fewer lower-grade tonne removes ₦600,000 and the higher conversion cost removes ₦800,000.
- THE DOCTRINE
This calculation does not establish whether the intervention pays after software, technical evaluation, additional working capital and support costs. Those costs belong in the programme case.
When a paper quotes the value of extra tonnes, ask what happened to the mix and the conversion cost.
12:00 – 17:00
Test the alternative; treat quality and safety as constraints
- A fair comparison asks what would have happened without the intervention. Feed quality, prices, maintenance condition and demand may have changed at the same time, so a before-and-after result does not by itself isolate the model's contribution.
- Where feasible, compare similar campaigns or batches under a documented evaluation plan. If the pilot combined better recordkeeping, staff training and recommendations, call it a combined intervention. Do not turn a short trial into an annual promise by multiplication alone.
- Commercial analysis does not grant permission to ship off-specification product or operate beyond an approved limit. Even the teaching example's permitted lower grade needs a real specification, suitable customer use and authorised disposition.
- IN THE PLANT · THE WORKED CASE
- Plant. In a fictional resin producer, the technical owner first establishes which campaign options are technically acceptable. An option outside that set is not an alternative, whatever its estimated margin.
IN THE MARGIN · THE PARALLEL- Margin. Finance then ranks only the feasible options and reconciles the bridge, while the technical owner explains the physical mechanism. If neither can reconstruct the bridge, the case is not ready.
The technical owner sets the feasible set; finance ranks within it, never the reverse.
17:00 – 22:00
Cash is a separate question, and scarce capacity has a price
- Contribution does not establish affordability. Material may be paid for before production while customer payment arrives long after delivery. Build a calendar of deposits, delivery payments, production expenditure, customer acceptance and expected collection; distinguish an invoice from cash in the bank; present the maximum cumulative funding need and the effect of a delayed receipt.
- The book's assumed capacity example: product P earns ₦120,000 contribution per accepted tonne and occupies a constrained vessel for two hours per tonne; Q earns ₦90,000 and occupies it for one hour. P earns ₦120,000 ÷ 2 = ₦60,000 per vessel-hour; Q earns ₦90,000 ÷ 1 = ₦90,000. Q is the better use when both are feasible and demand is available. The answer reverses when demand for Q is exhausted, another resource becomes limiting or P carries an omitted consequence.
- THE DOCTRINE
Contribution per tonne can conceal the resource that governs the period. The denominator must correspond to the actual constraint.
- A shadow price from an optimisation model is a local result, not a case for buying capacity. Ask for a sensitivity table around feed cost, accepted-product mix, constrained time and customer collection, without presenting the combined favourable case as a forecast.
A model that ignores the financing constraint can recommend work your company cannot execute.
22:00 – 25:00
Write it down, and Section 4
- Before you close the laptop, write the denominator your decision should be judged on and the resource that actually limits it.
- Draft the contribution bridge on one page: baseline and proposed quantities, accepted sales, relevant variable costs, programme costs and cash timing, with the assumptions that could reverse the decision. Separate recurring contribution from a one-time inventory cash release and from a deferred expenditure.
- Section 4 of your plan is The denominator, the alternative, the contribution bridge and the cash question.
Ask finance to reconcile it and the technical owner to explain it, in the same room.
Key Terms
Contribution
Revenue less the costs included in an expressly stated economic boundary; not a complete accounting profit measure.
Incremental cost
A cost that varies with the proposed decision within a stated boundary, unlike an allocation the choice leaves unchanged.
Funding trough
The largest cumulative cash deficit within a stated transaction boundary and timing model.
Shadow price
An optimisation model's estimated improvement from relaxing a constraint slightly, under its assumptions. A local analytical result.
Self-Check — Before You Submit
1. In the book's assumed campaign calculation, what is the incremental contribution of the technically approved intervention?
- A. ₦2 million
- B. ₦1.4 million
- C. ₦600,000
- D. ₦8.6 million
C. ₦600,000 — Contribution moves from ₦8 million to ₦8.6 million, an increase of ₦600,000. The ₦2 million value of two extra standard tonnes is the trap: it ignores the lost lower-grade tonne and the ₦800,000 rise in conversion cost.
2. A fictional formulator's team multiplies a short pilot's gain across a full year to state an annual benefit. What should you ask for?
- A. A result limited by suitable campaigns, capacity, demand and support cost, compared under a documented plan
- B. A larger multiplier for learning effects
- C. Approval now, with the figure revisited later
- D. A vendor letter confirming the figure
A. A result limited by suitable campaigns, capacity, demand and support cost, compared under a documented plan — The book says not to turn a short trial into an annual promise by multiplication alone. Approving now feels pragmatic, but it commits the company to a number nobody has bounded.
3. On the book's assumptions, a constrained vessel can make P (₦120,000 per tonne, two vessel-hours per tonne) or Q (₦90,000 per tonne, one vessel-hour). Both are feasible and in demand. Which uses the vessel better?
- A. P, because it earns more per tonne
- B. Q, at ₦90,000 per vessel-hour against ₦60,000 for P
- C. They are equal
- D. Neither, without a shadow price
B. Q, at ₦90,000 per vessel-hour against ₦60,000 for P — When the vessel-hour is the constraint, it is the denominator: ₦120,000 ÷ 2 gives ₦60,000 for P, ₦90,000 ÷ 1 gives ₦90,000 for Q. Contribution per tonne is the trap.
Today’s Assignment
For your decision, state the correct denominator, test the alternative, draw the contribution bridge and set out the cash question separately from margin.
- The denominator. The measure and units your decision should be judged on, the resource that actually limits it, and why gross or nominal measures would mislead in your business.
- The alternative. What would have happened without the intervention, the other changes that could explain a gain, and how you would compare under a documented plan.
- Constraints. Who in your company establishes the technically feasible set before finance ranks the options.
- The contribution bridge. Baseline and proposed quantities, accepted sales, variable costs, programme costs and cash timing on one page. Label any assumed values as assumed and show your arithmetic.
- The cash question. The calendar of commitments and receipts, the maximum cumulative funding need and the effect of a delayed receipt.
- The honest part. Which favourable movement your business has previously credited to the wrong cause. Four hundred to nine hundred words.
YOUR 90-DAY PLAN
Section 4 — The denominator, the alternative, the contribution bridge and the cash question.
How this is marked — the rubric
| Names the correct denominator tied to the actual binding constraint | 20 |
| Defines the alternative and the competing explanations, with a documented comparison | 20 |
| Builds a contribution bridge with labelled assumptions and correct arithmetic | 35 |
| Separates cash timing from margin and states the maximum funding need | 25 |
| Total · pass mark 70 | 100 |
Submitting session 4
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: 4, 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
Choosing the denominatorBuilding a contribution bridgeTesting the alternativeSeparating cash from marginPricing scarce capacity
Part Two · Feedstock and ProductionSESSION 5
Feedstock Quality and Procurement
Why This Matters
The purchase price is only one property of a feedstock decision. Composition, documented quality, compatibility with the process, delivery certainty and the cost of using the material all decide whether a low quotation becomes an expensive operating choice. A specialist supplier carries the same exposure without converting anything: it promised a product and must preserve its identity and evidence. Having connected your decision to contribution and cash in Session 4, today you look at what enters your business and when it is genuinely usable. A delayed campaign or a disputed lot can begin here, in a record that described the wrong date, the wrong denominator or the wrong lot.
25-Minute Module
0:00 – 5:00
Define the material before comparing the offer
- A low quotation can become an expensive operating choice if it creates rework, reduces accepted output or delays a committed order. For a supplier, a familiar trade name is not a substitute for a controlled specification.
- The procurement record identifies the material, manufacturer where relevant, grade, specification version, packaging, quantity basis and required documents. A kilogram of formulated product differs from a kilogram of active content, and a liquid quoted by volume needs a relevant density and reference basis before comparison with a mass quotation.
- Technical acceptance criteria come from the competent owner. Purchasing compares offers against them; it should not invent equivalence to make the comparison easier. Record which properties are mandatory, informative or need further evaluation. An unknown property remains an unknown.
- THE DOCTRINE
An AI assistant can extract fields from supplier documents and flag differences. Require it to show the document and location behind each field. A number extracted from a general brochure may not describe the offered lot.
If the record does not name the grade and specification version, the comparison has not started.
5:00 – 10:00
The assumed active-content comparison
- Plain language first. Active content is the part of each delivered kilogram that is the stated active component. Comparing on that basis prices the stated active component rather than the delivered mass; it does not establish technical equivalence.
Product (assumed) Price per kg delivered Accepted active mass fraction Price per kg of active content A ₦800 30 percent ₦800 ÷ 0.30 ≈ ₦2,666.67 B ₦900 35 percent ₦900 ÷ 0.35 ≈ ₦2,571.43 - The book labels these fictional inputs for arithmetic, not product recommendations, and assumes both products are already technically qualified and comparable. On this narrow basis B has the lower active-content price despite its higher price per delivered kilogram.
- THE DOCTRINE
The active-content calculation is valid only after the team has established that this denominator is relevant. It should never be used to declare two different chemistries interchangeable.
- What can still change the answer: carrier composition, impurities, compatibility, handling requirements, technical effectiveness, transport and approved use.
A cheaper kilogram and a cheaper active kilogram are different claims.
10:00 – 15:00
Preserve lot evidence, and treat variability as an operating question
- A supplier certificate describes the stated lot and test basis. Preserve it alongside your incoming inspection and laboratory results rather than overwriting it; where they disagree, maintain the exception and investigation record.
- Lot identity must survive receiving, storage, transfer, blending and dispatch. Repacking does not give a product a new origin. Where material is combined, the genealogy shows the contributing lots and quantities.
- Expose coverage before ranking suppliers. A vendor that supplies few lots may show no failures because the sample is small; one serving difficult applications may have more exceptions for reasons other than poor manufacture.
- IN THE PLANT · THE WORKED CASE
- Plant. In a fictional batch formulator, an association appears between one material source and off-grade batches. The useful dataset links feed lot to batch, approved conditions, output properties and disposition, and keeps alternative explanations: storage history, test-method changes and differences in the products made.
IN THE MARGIN · THE PARALLEL- Margin. Purchasing wants to drop the supplier. The association justifies targeted investigation, not a recipe change or supplier exclusion. Establish technical eligibility first, then compare price and lead time within the eligible set.
A supplier name is a weak substitute for measured material properties.
15:00 – 22:00
Import readiness, the uncertainty that matters and the usable date
- Required approvals, documents, inspections and release steps belong in the lead-time model. NAFDAC's guidelines for obtaining a permit to import industrial and laboratory chemicals (new applicants) request chemical identity, quantities, intended uses, relevant composition and HS codes, and manufacturer safety information. The PDF the book reviewed displays a review date of 16 November 2023 and a future date labelled effective date of 15 November 2028; the book says those labels need clarification against the Agency's operative guidance before determining a current obligation.
- SON's published SONCAP FAQ distinguishes product certification from consignment certification and describes a qualified exemption concerning chemicals used by bona fide manufacturers for food and medicines; it is not a blanket exemption for raw-material chemicals. In July 2025 a Nigeria Customs Service statement published by the Federal Ministry of Information described a phased Form M migration from NICIS II to B'Odogwu and the associated PAAR workflow, evidence that trade-system arrangements can change.
- THE DOCTRINE
A forecasting application can identify missing documents or an approaching deadline. It cannot establish regulatory compliance by generating a completed-looking checklist.
- An average lead time can conceal a long tail of delays. For a committed campaign, management needs the probability and consequence of material missing its required date, and the feasible responses: delaying the campaign, buying an approved alternative, using capacity for another product or obtaining a customer-approved revised date. A technical substitution enters change control.
- Separate promised shipment, actual shipment, arrival, documentary completion, customs release, site receipt and technical release. Arrival at a port cannot stand in for technically released material. Keep open shipments in the history; where evidence is sparse, use explicit delay scenarios rather than a precise probability.
Forecast the event the campaign needs, not the event that is easiest to record.
22:00 – 25:00
Write it down, Section 5, and Practice Exam 1
- Before you close the laptop, write the date your decision actually depends on: not arrival, but the usable date. If your business runs no plant, apply this to the product you promise a customer, or to the counterpart you depend on.
- Sketch an obligation register linked to your procurement plan: source, competent reviewer, required evidence and current status.
- Section 5 of your plan is The material defined, lot evidence preserved and the usable date modelled.
- Practice Exam 1 opens now: fifteen questions, closed book, covering Sessions 1 to 5. It closes Week 1.
Keep one reconciled identity from quotation to physical lot.
Key Terms
Active-content basis
A quantity comparison using the stated active component and denominator. It does not establish technical equivalence.
Lot
An identifiable quantity of material whose identity should remain traceable through movement and use.
Batch genealogy
The record linking a batch to the contributing material lots, quantities and relevant processing history.
Obligation register
A register linked to the procurement plan stating each obligation's source, competent reviewer, required evidence and status.
Self-Check — Before You Submit
1. In the book's assumed comparison, product B costs ₦900 per kilogram with an accepted active mass fraction of 35 percent. What is its approximate price per kilogram of active content?
- A. ₦315.00
- B. ₦2,666.67
- C. ₦900, the delivered price
- D. ₦2,571.43
D. ₦2,571.43 — ₦900 divided by 0.35 is approximately ₦2,571.43. ₦2,666.67 is product A's result; ₦900 ignores the carrier; multiplying instead of dividing gives ₦315.
2. A fictional specialist supplier's planning screen shows the expected port arrival date for an imported lot that a customer order depends on. What should the forecast predict?
- A. The date the material is technically released and usable
- B. The supplier-promised shipment date
- C. The port arrival date
- D. The average lead time across past consignments
A. The date the material is technically released and usable — A forecast should predict the event the order needs; arrival at a port cannot stand in for technically released material. An average lead time is tempting but can conceal a long tail of delays.
3. A fictional formulator's supplier scoring tool rates a material highly despite one failed mandatory property, because its price and lead time are excellent. What does the book require?
- A. Accept the overall score
- B. Lower the weight on the failed property
- C. Establish technical eligibility first, then compare economics within the eligible set
- D. Let purchasing decide whether the property matters
C. Establish technical eligibility first, then compare economics within the eligible set — The book rejects a composite score that compensates for a failed mandatory property with price or lead time. Balancing sounds fair, but eligibility comes first, and acceptance criteria come from the competent technical owner.
Today’s Assignment
For the material or product your decision depends on, define it before comparing offers, set how lot evidence is preserved, treat import readiness as part of availability and model the usable date with evidence.
- The material defined. Identity, grade, specification version, packaging, quantity basis and required documents; mandatory, informative and unknown properties; and who in your company sets the acceptance criteria.
- Lot evidence. How supplier certificates and your own results are kept side by side, how lot identity survives each movement and how disagreements are recorded.
- Import readiness. The approvals, documents and release steps that apply, who reviews them, and what your business has not yet verified against current official sources.
- The usable date. The events you will separate, the feasible responses if material is late, and an explicit delay scenario with its consequence for accepted dispatch.
- If your business has no plant, apply each point to the product you supply or the counterpart you depend on.
- The honest part. Where your company treats a promise, an arrival or a brochure as evidence. Four hundred to nine hundred words.
YOUR 90-DAY PLAN
Section 5 — The material defined, lot evidence preserved and the usable date modelled.
How this is marked — the rubric
| Defines the material with specification version, quantity basis and mandatory properties set by a competent owner | 25 |
| Preserves supplier and company results, lot identity and genealogy, and records disagreements | 25 |
| Treats import readiness as part of availability, with named reviewers and unverified obligations stated | 20 |
| Models the usable date with separated events, feasible responses and an explicit delay scenario | 30 |
| Total · pass mark 70 | 100 |
Submitting session 5
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: 5, 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 1 — Leadership, Margin and Feedstock 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 1 →Closed book, no notes, and Coach Ritchie switches himself off while you sit it.Capabilities Covered
Defining the materialComparing on the right basisPreserving lot evidenceImport readiness as availabilityForecasting the usable date