Established models, and
where each one belongs

Start from why an organisation exists, and let what it does follow from that rather than the other way round.

The argument is that most organisations can say what they do and how they do it, and very few can say why — and that the ones people follow are the ones that lead with the why. It is less a method than an ordering principle, and its usefulness is in the discipline of refusing to answer the easy question first.

Reach for it when

When a purpose statement keeps coming out as a description of the product. Working outside in is the fastest way to find that you have written a what and called it a why.

Where it stops

It orders the question; it does not answer it. Nothing in it tells you whether the why you arrived at is true, or whether anyone inside the company would recognise it.

Simon Sinek, Start With Why, Portfolio, 2009.

The part of a company that is meant never to change, separated from everything that should.

Built on a six-year study of companies that had lasted, it splits the enduring core — values and a reason for being that could still be pursued in a hundred years — from the strategies and practices that ought to change constantly. The distinction is the point: most statements fail because they mix the two, and then either the values drift or the strategy calcifies.

Reach for it when

When drafting or testing a purpose, and when deciding what a company is not allowed to trade away.

Where it stops

It is descriptive of companies that survived, which is a survivor’s sample. It tells you what the long-lived had in common, not that having it will make you long-lived.

James C. Collins & Jerry I. Porras, Built to Last, HarperBusiness, 1994; and “Building Your Company’s Vision”, Harvard Business Review, 1996.

Five tests a written objective has to pass before it counts as one.

Specific, measurable, assignable, realistic, time-related. The letters have been rewritten many times since, but the original point survives all the versions: an objective nobody could prove was met is not an objective, it is a sentiment with a deadline attached to it later.

Reach for it when

As a last check before a goal is written down and agreed. It takes a minute and it catches most of what goes wrong.

Where it stops

It tests the wording, not the ambition. A goal can pass all five and still be the wrong goal, and the criteria quietly favour what is easy to measure.

George T. Doran, “There’s a S.M.A.R.T. Way to Write Management’s Goals and Objectives”, Management Review, 1981.

A qualitative objective paired with a few numbers that would have to move for it to be true.

The structure separates the thing you want, stated in plain language, from the evidence that it happened. Its real work is in the cadence around it — set for a quarter, reviewed openly, and deliberately set high enough that full attainment is not expected.

Reach for it when

When goals across a company need to line up without being dictated line by line from the top.

Where it stops

It is a system of attention, not of strategy. Tied to pay it stops being ambitious within a quarter, and it will make a company very good at whatever it happens to have written down.

Andrew S. Grove, High Output Management, Random House, 1983; John Doerr, Measure What Matters, Portfolio, 2018.

A single long-range goal clear enough to need no explanation and large enough to change the company on the way.

It is defined by a horizon of ten to thirty years, a finish line anyone can recognise, and a genuine risk of failure. The mechanism is commitment: a goal that cannot be quietly reinterpreted three years in has to be either pursued or abandoned in public.

Reach for it when

When long-term goals keep coming out as this year’s plan with a bigger number on it.

Where it stops

It rewards clarity over correctness. A company can commit to a vivid goal for twenty years and be pointed the wrong way for all of them.

James C. Collins & Jerry I. Porras, Built to Last, HarperBusiness, 1994.

Separating the business you are running from the one you are building and the one you are only betting on.

Horizon one defends and extends what already earns; horizon two builds what is emerging; horizon three places options on what might exist later. Each needs different funding, different measures and different people, and the model exists because judging all three by the first one’s standards kills the other two every time.

Reach for it when

When long-term goals and this quarter’s numbers are being argued in the same meeting with the same yardstick.

Where it stops

The horizons are a way of talking, not a portfolio theory. Nothing in it says how much belongs in each, and it is easily used to protect pet projects from any measurement at all.

Mehrdad Baghai, Stephen Coley & David White, The Alchemy of Growth, Perseus, 1999.

Translating a few company-level aims down through every level, and negotiating them back up before anyone commits.

Sometimes called policy deployment. What distinguishes it from cascading targets is the second movement: each level proposes what it can actually do, and the disagreement between that and what was asked for is resolved explicitly rather than absorbed silently.

Reach for it when

When strategy is agreed at the top and quietly renegotiated by everyone below it as the year goes on.

Where it stops

It is slow, and it assumes a stable enough year that a plan negotiated in January still describes reality in June.

Yoji Akao (ed.), Hoshin Kanri: Policy Deployment for Successful TQM, Productivity Press, 1991.

A loop that treats every change as a small experiment with a result you are obliged to look at.

Plan the change and what you expect, make it small, check what actually happened against what you predicted, then either standardise it or start again. The discipline that matters is writing the prediction down first, which is the step most often skipped.

Reach for it when

For tactical goals, where the point is a short cycle and a real answer rather than a defended forecast.

Where it stops

It improves what exists. A loop of small corrections will refine the wrong thing indefinitely without ever raising the question.

Walter A. Shewhart, Statistical Method from the Viewpoint of Quality Control, 1939; W. Edwards Deming, Out of the Crisis, MIT Press, 1986.

Managers and their people agree objectives together, and performance is judged against those rather than against activity.

Drucker’s argument was that an organisation only pulls in one direction if each part knows what it is contributing and has agreed to it, rather than been handed it. Nearly every goal system since — including objectives and key results — descends from this, and so do its failure modes, which Drucker himself warned about.

Reach for it when

When people are being measured on how busy they look rather than on anything that was agreed in advance.

Where it stops

Applied mechanically it becomes a paperwork exercise that rewards easily agreed objectives. Drucker later said the method works only where the strategy behind the objectives is sound, which is the harder half.

Peter F. Drucker, The Practice of Management, Harper & Row, 1954.

A strategy is a diagnosis, a guiding policy and a set of coherent actions — and anything without all three is a wish list.

The diagnosis names what is actually in the way. The guiding policy is the approach chosen to deal with it. The actions are coordinated steps that follow from it and support each other. The argument is that most documents called strategy contain only ambitions, and that the absence of a diagnosis is what makes them impossible to argue with.

Reach for it when

As a test on any strategic goal: can you state the obstacle it answers, and does anything follow from it that rules something else out?

Where it stops

It is a standard for what a strategy must contain, not a method for finding one. It can tell you that yours is empty without telling you what to put in it.

Richard P. Rumelt, Good Strategy / Bad Strategy, Crown Business, 2011.

Four directions growth can come from, ordered by how much is unfamiliar in each.

Selling more of what you have to who you already sell to; taking what you have to new markets; building something new for the customers you know; or doing both at once. The order is the argument — each step away from the known multiplies what has to be learned, and diversification is the one that fails most.

Reach for it when

When strategic goals need to be laid out side by side and someone has to say plainly how much new ground each one crosses.

Where it stops

It sorts directions; it does not size them. Two options in the same box can carry entirely different risk, and the matrix has nothing to say about that.

H. Igor Ansoff, “Strategies for Diversification”, Harvard Business Review, 1957.

There are three ways to win a market, and trying to do more than one at a time is how companies end up winning none.

Compete on cost, compete on being meaningfully different, or serve a narrow segment better than anyone serving everyone. The famous warning is what makes it a strategy tool rather than a menu: a company that pursues cost and differentiation at once, without being clearly best at either, ends up in what Porter called stuck in the middle — carrying the cost of both and the advantage of neither.

Reach for it when

When strategic goals commit to being cheaper and better and more specialised, all in the same document.

Where it stops

The claim that the three cannot be combined has been argued about for forty years, and several companies have plainly managed it. Read it as a demand to say which one you are betting on, not as a law.

Michael E. Porter, Competitive Strategy, Free Press, 1980.

Teams pass through forming, storming, norming and performing, and the difficult stage is not a failure of the team.

Drawn from a review of studies of small groups, it describes a sequence rather than prescribing one. Its practical value is almost entirely in the second stage: knowing that conflict over roles and direction is a normal passage stops managers from treating it as evidence that they hired the wrong people.

Reach for it when

When a new team is struggling and someone is about to reorganise it.

Where it stops

Groups do not move through it neatly, and they go backwards when membership changes. It is a vocabulary for what is happening, not a schedule.

Bruce W. Tuckman, “Developmental Sequence in Small Groups”, Psychological Bulletin, 1965.

People contribute to a team in recognisable ways, and a team of one kind fails in a predictable direction.

Built from years of business-game experiments at Henley, it separates the job someone holds from the role they play — the one who finishes things, the one who brings ideas, the one who asks the awkward question. The finding that made it useful was that teams of the most able people performed badly, because they all played the same role.

Reach for it when

When composing a team, or when trying to name what a team that keeps stalling is actually missing.

Where it stops

Self-reported role profiles are not stable traits, and the psychometrics have been contested for decades. Use it as a way of talking about a gap, not as a test of a person.

R. Meredith Belbin, Management Teams: Why They Succeed or Fail, Heinemann, 1981.

What makes people dissatisfied at work and what motivates them are two different lists, not two ends of one.

Pay, conditions, policy and supervision are hygiene factors: get them wrong and people are unhappy, get them right and they are merely not unhappy. Achievement, recognition, responsibility and the work itself are what actually motivate. The consequence is uncomfortable and often ignored — you cannot fix a motivation problem by improving the conditions.

Reach for it when

When engagement is falling and the proposed answer is a benefit, a bonus or an office refurbishment.

Where it stops

The original method asked people to recall good and bad episodes, and people attribute the good to themselves and the bad to circumstance. The split is cleaner in the model than in the evidence.

Frederick Herzberg, Bernard Mausner & Barbara Snyderman, The Motivation to Work, Wiley, 1959.

People attend to basic needs before higher ones, and what motivates someone depends on which are already met.

Physiological needs, safety, belonging, esteem and self-actualisation, each becoming pressing as the one below it is satisfied. Its practical use at work is narrow but real: an offer of purpose and growth lands very differently on someone who is worried about whether the job still exists in six months.

Reach for it when

When engagement work reaches for meaning and mastery while pay, workload or job security are unresolved.

Where it stops

The pyramid was not Maslow’s and the strict ordering has not held up in research — people pursue several levels at once, and cultures order them differently. It is a useful reminder, not a sequence to manage by.

Abraham H. Maslow, “A Theory of Human Motivation”, Psychological Review, 1943.

Eight steps for changing how an organisation works, beginning with urgency and ending with making the change stick.

Drawn from watching change efforts fail, the sequence runs from establishing urgency and building a guiding coalition, through a clear vision and communicating it relentlessly, to short-term wins and anchoring the change in the culture. Kotter’s finding was that skipping any step feels faster and reliably costs more later; most efforts fail at the first, because everyone is already busy.

Reach for it when

Before a change programme starts, and when one has stalled and nobody can say at which step.

Where it stops

It is top-down and linear, and it was drawn from large corporations in the 1990s. Anything genuinely emergent will not wait for step three.

John P. Kotter, “Leading Change: Why Transformation Efforts Fail”, Harvard Business Review, 1995; Leading Change, HBS Press, 1996.

There is no single right way to lead — the right amount of direction depends on how ready the person is for that particular task.

Four styles, from telling someone exactly what to do, through coaching and supporting, to handing the task over entirely. The variable is not the person’s worth but their competence and confidence at the task in front of them, which means the same person needs different handling on two different jobs.

Reach for it when

When a capable person is being micromanaged, or a new one is being left alone and called a poor fit.

Where it stops

Judging someone’s readiness is a subjective call, and the empirical support is weaker than the model’s popularity suggests. It is a way of checking your instinct, not a measurement.

Paul Hersey & Kenneth H. Blanchard, “Life Cycle Theory of Leadership”, Training and Development Journal, 1969.

People do not buy products; they hire something to make progress in a situation.

The unit of analysis is the job and the circumstance, not the customer and their demographics. It reframes competition too: whatever else someone could hire for the same job is a competitor, however different it looks. The famous milkshake study is not the theory — it is an illustration of asking what someone was trying to get done at seven in the morning.

Reach for it when

When a value proposition is written as a list of features, or when segmentation by who people are has stopped explaining who buys.

Where it stops

Jobs are inferred from interviews and are easy to write to fit the product you already have. Nothing in the method stops you from discovering the job you hoped for.

Clayton M. Christensen, Scott Cook & Taddy Hall, “Marketing Malpractice”, Harvard Business Review, 2005; Competing Against Luck, HarperBusiness, 2016.

Some features delight, some merely satisfy in proportion, and some are only noticed when they are missing.

It sorts what a product offers into categories with different relationships to satisfaction — the expected, where more brings no credit and absence brings anger; the linear, where more is straightforwardly better; and the unexpected, which delights precisely because nobody asked. Its sharpest observation is that categories decay: today’s delight becomes tomorrow’s expectation.

Reach for it when

When deciding what to build next and every candidate feature is being argued as equally important.

Where it stops

Placing a feature in a category depends on a survey with a particular question format, and the answers move with market and segment. It is a snapshot, not a property of the feature.

Noriaki Kano et al., “Attractive Quality and Must-Be Quality”, Journal of the Japanese Society for Quality Control, 1984.

A structured way of checking that what you offer answers something the customer actually struggles with.

It sets what customers are trying to do, what gets in the way and what they hope for on one side, and what the product does about each on the other, then asks where the two sides genuinely meet. Its work is in the honesty of the gaps — the pains nothing addresses, and the features nobody was in pain about.

Reach for it when

When a proposition has been written from the inside out and needs testing against a real customer situation.

Where it stops

It organises what you believe about the customer. Filled in from a meeting room it will feel rigorous and be entirely wrong.

Alexander Osterwalder, Yves Pigneur, Gregory Bernarda & Alan Smith, Value Proposition Design, Wiley, 2014.

Treat a new offer as a set of assumptions to be tested cheaply, rather than a plan to be executed.

Build the smallest thing that tests the riskiest assumption, measure what people actually do with it, learn, and either persevere or change direction. Its contribution was to name what was being risked: not money, but the time spent building something carefully before finding out whether anyone wanted it.

Reach for it when

Whenever a value proposition is being built out before anyone has paid for a version of it.

Where it stops

It suits things that can be released in pieces and measured quickly. Applied to work with long cycles, safety consequences or heavy fixed costs, minimum and viable start to fight each other.

Eric Ries, The Lean Startup, Crown Business, 2011; building on Steve Blank’s customer development.

Understand the person’s situation before defining the problem, and prototype the answer before committing to it.

Empathise, define, ideate, prototype, test — with the loop run more than once. What separates it from ordinary product work is the insistence that the problem statement is itself an output, arrived at after watching people rather than before. Most of its value shows up in the first two steps, which are the ones under pressure to be skipped.

Reach for it when

When a proposition is being designed from a solution someone already has in mind.

Where it stops

Widely taught and loosely practised — running a workshop with sticky notes is not the method. It also says nothing about whether the resulting idea is commercially viable.

Tim Brown, “Design Thinking”, Harvard Business Review, 2008; developed at IDEO and the Hasso Plattner Institute of Design at Stanford.

Nine linked areas that together describe how an organisation creates, delivers and captures value.

Its contribution was making a business model a single visible object that a group could argue over, instead of thirty pages nobody read. Because the areas are linked, changing one obliges you to look at the others — which is where most of the value is, and where most people stop.

Reach for it when

Early, when the shape of the business is still being decided, and whenever a change to pricing or channel needs its consequences traced.

Where it stops

It describes; it does not evaluate. A coherent-looking model can still have no customer, and the canvas has nothing to say about competition or timing.

Alexander Osterwalder & Yves Pigneur, Business Model Generation, Wiley, 2010.

Four questions about price that between them mark out the range a market will tolerate.

Rather than asking what someone would pay, it asks at what price something would seem too expensive, expensive but worth considering, cheap, and so cheap that quality is in doubt. The four answers plotted together give a band with a floor and a ceiling, and the floor is the finding people forget: too low reads as a warning.

Reach for it when

When setting a price for something new, where there is no history to extrapolate from.

Where it stops

It measures stated intent, not behaviour, and it says nothing about volume at any price. It gives you a plausible range to test, not a price.

Peter H. van Westendorp, “NSS Price Sensitivity Meter”, ESOMAR Congress, 1976.

Set price from the value delivered relative to the customer’s next best alternative, not from what it cost to make.

The method works from a reference price — what the buyer would otherwise do — and adds what your offer is worth more, minus what it is worth less. Its discipline is that both halves must be quantified from the buyer’s position, which is precisely the work that cost-plus pricing exists to avoid.

Reach for it when

When margins are being set by adding a percentage to cost, or when a discount is about to be given without knowing what was given away.

Where it stops

It needs a defensible estimate of what the alternative is worth to someone else, and that estimate is often unavailable and always arguable.

Thomas T. Nagle & Reed K. Holden, The Strategy and Tactics of Pricing, Prentice Hall, 1987.

The few things a company knows how to do that open more than one market and are hard to copy.

A competence qualifies on three tests: it gives access to a range of markets, it makes a real difference to what the customer perceives, and it is difficult for competitors to imitate. The argument behind it is that companies compete on capabilities long before they compete on products, and that treating business units as independent hides them.

Reach for it when

When deciding what to keep inside the company and what can safely be bought or outsourced.

Where it stops

Almost anything can be argued into the definition after the fact. Applied to yourself it flatters; the tests are only useful if someone is allowed to fail them.

C. K. Prahalad & Gary Hamel, “The Core Competence of the Corporation”, Harvard Business Review, 1990.

Four questions that decide whether something a company has is an advantage or just an expense.

Is it valuable, is it rare, is it costly to imitate, and is the organisation arranged to exploit it. Only what passes all four supports advantage that lasts; passing the first two gives you parity or a temporary lead. The fourth question is the one companies fail — holding something genuinely rare and being organised in a way that wastes it.

Reach for it when

As a sorting test across a list of claimed strengths, when everything on the list has been called a strength by the person who owns it.

Where it stops

The answers are judgements, and imitability in particular can only really be assessed backwards. It structures an argument rather than settling one.

Jay B. Barney, “Firm Resources and Sustained Competitive Advantage”, Journal of Management, 1991.

Look at a company as a bundle of resources rather than a set of products, and advantage comes from what it holds.

The move is to analyse the firm from the resource side — what it owns, knows and controls — instead of from the market side. It matters for assets because it takes seriously the things that never appear on a balance sheet: a customer base, an accumulated dataset, a reputation, a way of working that took ten years to build.

Reach for it when

When taking stock of what a business actually has, particularly the parts nobody thought to write down.

Where it stops

It is a way of seeing rather than a procedure, and it is weak on change — a resource that carried a company for a decade can be worth nothing the year the market turns.

Birger Wernerfelt, “A Resource-Based View of the Firm”, Strategic Management Journal, 1984.

Sort what the business owns by how fast its market is growing and how strong its position in it is, and fund each kind differently.

Four boxes, and the names have outlived the theory: strong positions in growing markets need investment, strong positions in slow markets pay for it, weak positions in growing markets are a bet, and weak positions in slow markets are a decision nobody has taken. Its real subject is cash — which parts of a portfolio generate it and which consume it.

Reach for it when

When a business has several products, brands or units and they are all being funded as though they were the same kind of thing.

Where it stops

Market share and market growth are crude proxies for a position’s value, and the boxes ignore how units support one another. Used literally it has closed businesses that were quietly holding something together.

Bruce D. Henderson, “The Product Portfolio”, Perspectives, Boston Consulting Group, 1970.

Every system has one bottleneck at a time, and improving anything else changes nothing.

Find the constraint, get everything possible out of it, subordinate the rest of the system to it, then lift it — and when it moves, start again. The counter-intuitive part is subordination: deliberately running other parts below capacity, which looks like waste on every local measure and is the only thing that raises throughput.

Reach for it when

When a process has been optimised in many places and the output has not moved.

Where it stops

It assumes a system with a stable, findable bottleneck. In knowledge work the constraint moves, is often a person’s attention, and does not sit still to be measured.

Eliyahu M. Goldratt & Jeff Cox, The Goal, North River Press, 1984.

Define value from the customer’s side, then remove everything in the flow that does not create it.

Five principles: specify value, map the stream that delivers it, make it flow, let the customer pull, and pursue perfection. Underneath is a definition worth stealing on its own — waste is anything the customer would not knowingly pay for, which includes most waiting, most rework and most of what is done just in case.

Reach for it when

When work moves slowly through a process and nobody can say where the time goes.

Where it stops

Removing slack removes resilience with it. A system tuned for flow is fragile to variation, as several lean supply chains discovered at the same moment.

James P. Womack & Daniel T. Jones, Lean Thinking, Simon & Schuster, 1996; Taiichi Ohno, Toyota Production System, 1978.

Break the business into the activities it actually performs, and ask which of them the customer is paying for.

Primary activities — bringing things in, making them, getting them out, selling them, supporting them — run across the middle, with procurement, technology, people and infrastructure supporting all of them. The point is not the diagram but the question it forces at every box: does this activity add value the customer would pay for, does it cost less than a competitor’s, and if neither, why is it here?

Reach for it when

When looking for where margin actually comes from, and when deciding what to keep, improve or hand to someone else.

Where it stops

It was drawn for manufacturing and fits a linear flow best. Service and software businesses often have to redraw it before it says anything useful.

Michael E. Porter, Competitive Advantage, Free Press, 1985.

Seven things that have to agree with one another for an organisation to work: strategy, structure, systems, shared values, style, staff and skills.

Three of them are hard and can be changed by decision; four are soft and change slowly whatever anyone decides. The argument is that a change to one obliges a change to the others, which is why a new strategy imposed on an unchanged structure, unchanged incentives and unchanged habits reliably comes to nothing.

Reach for it when

Before a reorganisation, and when a strategy has been announced twice and nothing downstream has moved.

Where it stops

It says what has to agree, not what any of them should be. As a diagnosis it is strong; as a design it is empty.

Robert H. Waterman Jr., Thomas J. Peters & Julien R. Phillips, “Structure Is Not Organization”, Business Horizons, 1980.

The players around a business are customers, suppliers, competitors and complementors — and the last one is usually missing from the map.

A complementor is anyone who makes your offer more valuable by existing. Adding them to the picture changes what a relationship is for: the same company can be a competitor in one respect and a complementor in another, and treating that as a contradiction leaves value unclaimed on both sides.

Reach for it when

When mapping who a partnership actually serves, or when a competitor turns out to be growing the market you both sell into.

Where it stops

It draws the board. Who captures the value created is a bargaining question the model deliberately leaves open.

Adam M. Brandenburger & Barry J. Nalebuff, Co-opetition, Currency Doubleday, 1996.

Whether to do something yourself or have someone else do it turns on the cost of the arrangement, not the price of the work.

Searching, negotiating, specifying, monitoring and enforcing all cost something, and they rise sharply when what is being exchanged is hard to specify or requires investment useful to only one buyer. That is the real argument for keeping something in-house: not that it is cheaper to make, but that the contract would be impossible to write.

Reach for it when

At every make-or-buy decision, and when a partnership keeps needing renegotiation.

Where it stops

The costs it turns on are largely unmeasurable in advance. It explains arrangements convincingly after the fact and predicts them poorly.

Oliver E. Williamson, Markets and Hierarchies, Free Press, 1975; The Economic Institutions of Capitalism, Free Press, 1985.

A company is answerable to everyone who can affect it or is affected by it, not to shareholders alone.

Its practical core is an instruction to identify those groups explicitly and to treat their interests as inputs to strategy rather than as constraints discovered later. The argument is not primarily ethical: relationships that are managed produce better information and fewer surprises than relationships that are merely endured.

Reach for it when

Whenever a decision’s consequences reach past the people in the room.

Where it stops

It gives no rule for what to do when interests genuinely conflict, which is exactly when the question gets hard.

R. Edward Freeman, Strategic Management: A Stakeholder Approach, Pitman, 1984.

Place each stakeholder by how much power they hold and how much they care, and how to treat them follows.

High power and high interest means manage closely; high power and low interest means keep satisfied, because they will act if they start caring; low power and high interest means keep informed; the rest, monitor. The value is in the second box, which is where unpleasant surprises come from.

Reach for it when

Before a change that will affect several groups differently, and before deciding who is consulted rather than told.

Where it stops

Both axes are estimates, and both move. A grid drawn once and filed is a description of last year.

Aubrey L. Mendelow, “Environmental Scanning: The Impact of the Stakeholder Concept”, ICIS Proceedings, 1981.

Break return on equity into margin, asset turnover and leverage, so the number says where it came from.

The same return can be produced by a high margin on few sales, a thin margin on many, or borrowed money — and those are three entirely different businesses with three different risks. Decomposition turns a headline ratio into a diagnosis, and it is one of the oldest management tools still in daily use.

Reach for it when

When a return figure is being compared across periods or against a competitor and the comparison is doing no work.

Where it stops

It works on reported accounting numbers, with everything that implies about timing, policy and judgement. It explains the arithmetic, not the business.

Developed at E. I. du Pont de Nemours from 1912; see Alfred D. Chandler, Strategy and Structure, MIT Press, 1962.

How much has to be sold before the fixed costs are covered, and what each sale after that is worth.

Separating fixed from variable cost gives a contribution per unit and a volume at which the business stops losing money. The insight that survives beyond the arithmetic is about shape: a high-fixed-cost business is fragile below the line and very profitable above it, and knowing which kind you are running changes what risk you can take.

Reach for it when

Before committing to fixed costs, and when pricing something whose volume is uncertain.

Where it stops

It assumes costs stay neatly fixed or variable and that price holds as volume grows. Both assumptions fail exactly where the decision is interesting.

Walter Rautenstrauch, The Economics of Business Enterprise, Wiley, 1939; standard in management accounting since.

What it costs to win one customer, set against what that customer is worth over the whole relationship.

Acquisition cost against lifetime value, with the payback period between them. Its discipline is that growth stops being self-evidently good: a business that spends more to win a customer than the customer will ever return simply loses money faster the more it grows, and revenue charts hide that for years.

Reach for it when

Whenever growth is being funded, and before increasing spend on any channel.

Where it stops

Lifetime value is a forecast wearing the clothes of a measurement. Change the assumed retention slightly and the whole case turns over.

Developed in direct marketing and subscription practice; see David Skok, “SaaS Metrics 2.0”, 2013, for the standard treatment.

Sort what you buy by how much it matters and how risky it is to get, and buy each kind differently.

Four categories: the routine, the leveraged, the bottleneck and the strategic. What it changed was the assumption that purchasing is one activity — a bottleneck item worth very little can stop a factory, and treating it like stationery because it is cheap is how that happens.

Reach for it when

When supplier relationships are being managed uniformly, or when a trivial component has just halted production.

Where it stops

It sorts categories, not individual relationships, and it says nothing about what the supplier thinks of you — who may have their own matrix with you in the routine box.

Peter Kraljic, “Purchasing Must Become Supply Management”, Harvard Business Review, 1983.

Small changes in end demand grow into large swings the further back up the chain you go.

Each link reacts to the orders in front of it rather than to real demand, adds a margin of safety, and passes an amplified signal backwards. The important finding is that this happens with entirely rational behaviour at every step — nobody has to be foolish for the chain to end up wildly over- and under-stocked in turn.

Reach for it when

When stock swings between shortage and glut, and when deciding who in a chain gets to see actual demand.

Where it stops

Naming it does not damp it. The remedies — shared information, shorter lead times, less order batching — need agreement between parties who each benefit from their own buffer.

Jay W. Forrester, Industrial Dynamics, MIT Press, 1961; Hau L. Lee, V. Padmanabhan & Seungjin Whang, “The Bullwhip Effect in Supply Chains”, Sloan Management Review, 1997.

A shared vocabulary and set of measures for the standard processes every supply chain performs.

It names the chain’s activities — plan, source, make, deliver, return, enable — at increasing levels of detail and attaches standard metrics to each. Its worth is comparability: without a shared definition, two companies measuring the same thing are measuring different things, and neither knows it.

Reach for it when

When comparing performance against anyone outside the company, or when two organisations have to describe a shared chain to each other.

Where it stops

It standardises description, not design. A chain can score well on every standard measure and still be the wrong chain for the strategy.

Supply Chain Operations Reference model, Supply Chain Council, from 1996; now maintained by ASCM.

One number combining how much of the time equipment runs, how fast, and how much of what it makes is good.

Availability times performance times quality. Multiplying rather than averaging is the design decision that makes it honest: three respectable-looking figures produce a poor one, which is usually a truer description of the line than any of them alone.

Reach for it when

When a line is believed to be running well and the output does not agree.

Where it stops

It measures the equipment, not the value. A machine can run at high effectiveness producing stock nobody ordered, and the number will approve.

Seiichi Nakajima, Introduction to TPM, Japan Institute of Plant Maintenance, 1982.

Draw the whole path a product takes, with the waiting included, and the waiting turns out to be most of it.

Both material and information flow are drawn on one page, with processing time and elapsed time recorded separately. The ratio between them is the finding: in most processes the time spent actually adding value is a small fraction of the time the thing exists, and nobody believes that until they have drawn it.

Reach for it when

Before improving any step, so the effort goes where the time actually is.

Where it stops

It is a snapshot of a repeatable flow. Applied to work that varies from case to case it produces a tidy picture of something that never happens twice.

Mike Rother & John Shook, Learning to See, Lean Enterprise Institute, 1999.

Reduce variation until defects are rare, working through a fixed five-step cycle with the measurements agreed up front.

Define the problem, measure what is actually happening, analyse the causes, improve the process and control it so it does not drift back. The last step is the one that distinguishes it: most improvement efforts stop at improve, and the gain quietly disappears over the following year.

Reach for it when

Where a process runs often enough to be measured and the cost of defects is real.

Where it stops

It improves processes that already exist and rewards consistency, which is why organisations that push it hardest often become slower to try anything new. It also needs data most small operations do not have.

Developed at Motorola by Bill Smith from 1986; adopted at General Electric from 1995.

Who owns a risk, who oversees it, and who independently checks both — kept as three distinct roles.

The first line owns and manages risk in the work itself; the second sets policy and monitors; the third gives independent assurance to the governing body. The point is separation: assurance given by the people who would have to report their own failure is not assurance, whatever it is called.

Reach for it when

When setting up how compliance is organised, and when it is unclear who is actually responsible for a control.

Where it stops

Read as an org chart it produces bureaucracy and a first line that believes risk is someone else’s department. The 2020 revision exists largely because of that.

Institute of Internal Auditors, position papers from 1999; The IIA’s Three Lines Model, 2020.

A common set of principles and a process for identifying, assessing, treating and monitoring risk.

It gives a vocabulary and a cycle rather than a checklist, and deliberately is not certifiable. Its most useful contribution to ordinary practice is the insistence that risk is defined against objectives — which means you cannot assess risk at all until someone has said what the organisation is trying to do.

Reach for it when

When building a risk process from nothing, or when everyone in the room means something different by the word risk.

Where it stops

It is a framework for process, not a source of judgement. It will not tell you what your risks are or how much of them to accept.

ISO 31000, Risk management — Guidelines, International Organization for Standardization, 2009; revised 2018.

Brand strength is built in order: who you are, what you offer, what people think and feel, and only then loyalty.

Four levels, each unavailable until the one below it holds — identity, meaning, response, relationship. It is useful against a promise because it locates one: a promise is made at the meaning level and is redeemed, or not, at the response level, and no amount of repetition moves it up a floor.

Reach for it when

When a brand promise is being written, or when loyalty is being pursued directly by companies whose customers cannot say what they stand for.

Where it stops

The order is cleaner in the diagram than in life, and every level is measured by asking people, with everything that implies.

Kevin Lane Keller, “Building Customer-Based Brand Equity”, Marketing Management, 2001.

Satisfaction is not what you delivered; it is what you delivered measured against what was expected.

The same experience satisfies one person and disappoints another because they arrived expecting different things — and a brand promise is precisely the instrument that sets the expectation. It follows that an overstated promise damages a good product, and that raising quality without touching the promise raises satisfaction more cheaply than raising both.

Reach for it when

When deciding how much to promise, and when a genuinely good product keeps producing disappointed customers.

Where it stops

Expectations are formed by far more than your marketing — competitors, category norms, the last unrelated thing that went well. You set part of the reference point, not all of it.

Richard L. Oliver, “A Cognitive Model of the Antecedents and Consequences of Satisfaction Decisions”, Journal of Marketing Research, 1980.

Identity has six facets, some projected by the company and some constructed by the people who buy.

It separates what the brand sends — its physical qualities, its personality, its cultural roots — from what the recipient builds: the relationship, the reflected image of the typical user, and how buying it makes them see themselves. The separation is the contribution, because the second half is where most brands are actually decided and is the half nobody controls.

Reach for it when

When identity work keeps producing adjectives, and when the brand a company describes and the one its customers describe do not match.

Where it stops

Six facets invite six pieces of filled-in prose. Completed for its own sake it documents a brand rather than deciding one.

Jean-Noël Kapferer, Strategic Brand Management, Kogan Page, 1992.

A core identity that holds still, an extended identity around it, and an explicit statement of what the brand offers.

It asks a brand to be considered as a product, an organisation, a person and a symbol in turn, then to separate the few elements that must never move from the many that can flex by market and campaign. That separation is what lets a brand be consistent without being rigid, which is the practical problem it exists to solve.

Reach for it when

When a brand has to work across several markets or product lines without fragmenting.

Where it stops

It is a planning structure of considerable size. Applied to a small business it produces more document than brand.

David A. Aaker, Building Strong Brands, Free Press, 1996.

The colours, shapes, sounds and characters that make a brand recognisable before its name is read.

Each asset is judged on two things: how many people link it to the brand, and how many link it to that brand alone. Something known but shared with a competitor is worse than useless, because the recognition it earns is spent on someone else. The strategic consequence is patience — assets are built by consistency over years and destroyed in a single rebrand.

Reach for it when

Before changing anything visual, and when deciding what must survive a redesign.

Where it stops

It measures recognition, not meaning. An asset can be perfectly distinctive and stand for nothing anyone wants.

Jenni Romaniuk & Byron Sharp, “Conceptualising and Measuring Brand Salience”, Marketing Theory, 2004.

What matters is not whether people like a brand but whether it comes to mind in the situations where it could be bought.

Built on decades of panel data, it argues that most categories are bought by many light buyers rather than a loyal few, and that growth comes from being thought of in more buying situations by more people. It is deliberately unromantic about brand love, and the evidence behind it is uncomfortably large.

Reach for it when

When perception research reports strong affection among a small group and the brand is not growing.

Where it stops

It generalises from repeat-purchase consumer categories. In markets with few buyers and long considered purchases, the pattern it rests on is a different one.

Byron Sharp, How Brands Grow, Oxford University Press, 2010.

Positioning happens in the mind of the audience, not in the product, and the mind has very little room.

The argument is that people hold a short ordered list per category and defend it against new information, so the task is to find an unoccupied position and take it — including, if necessary, by defining a narrower category you can lead. Its most quoted advice is negative: attacking a position someone else already holds is the most expensive campaign there is.

Reach for it when

When entering a market where someone is already the obvious answer.

Where it stops

Written before the internet, in a world of three television channels and scarce attention. The scarcity is still real; the mechanics of getting into a category have changed completely.

Al Ries & Jack Trout, Positioning: The Battle for Your Mind, McGraw-Hill, 1981.

Some attributes get you considered at all; only the rest make anyone choose you.

Points of parity are what the category requires — absent them you are not a candidate, and exceeding them wins nothing. Points of difference are the reasons to prefer you. The common failure is investing in parity attributes and reporting the improvement as differentiation.

Reach for it when

When a positioning statement lists advantages that every competitor could also claim.

Where it stops

Which attribute is which changes as a category matures. Today’s difference becomes tomorrow’s entry requirement, and the classification has to be redone.

Kevin Lane Keller, Strategic Brand Management, Prentice Hall, 1998.

The value a name adds, held in awareness, associations, perceived quality and loyalty.

It made the case that a brand is an asset with identifiable components rather than a vague good feeling, and that each component can be tracked over time. The consequence for management is that brand spending becomes investment or depreciation rather than a cost line, and can be argued about on those terms.

Reach for it when

When brand work has to be justified to people who read a balance sheet, and when deciding whether to extend a name to something new.

Where it stops

The components are measured by survey and do not aggregate into a number anyone would put in the accounts. It gives an argument, not a valuation.

David A. Aaker, Managing Brand Equity, Free Press, 1991.

A required set of analyses — legal, behavioural and financial — before any monetary brand value is claimed.

It does not prescribe one calculation. It requires that whichever is used rests on all three: what is actually owned and protected, how the brand changes behaviour, and what that behaviour is worth. Its usefulness is as a standard to hold a valuation against, since brand valuations vary by multiples depending on who commissioned them.

Reach for it when

When a number is needed for a transaction, a licence or a balance sheet, and when someone else’s valuation needs testing.

Where it stops

Being methodologically sound does not make the assumptions right. Two compliant valuations of the same brand can still differ enormously.

ISO 10668, Brand valuation — Requirements for monetary brand valuation, International Organization for Standardization, 2010.

A small set of recurring character patterns that give a brand a consistent voice and role in a story.

The claim is that a limited number of figures — the ruler, the outlaw, the caregiver and so on — recur across cultures, and that a brand which occupies one consistently is easier to recognise and to write for. Its practical use is editorial: an archetype settles arguments about tone before they are had.

Reach for it when

When a brand’s voice differs by whoever last wrote the copy.

Where it stops

The Jungian foundation is contested and the categories are broad enough to fit anything. Treat it as a writing constraint rather than a psychological finding.

Margaret Mark & Carol S. Pearson, The Hero and the Outlaw, McGraw-Hill, 2001.

The shape of the story a culture keeps telling: departure, ordeal, return with something gained.

A comparative study of myth rather than a marketing method, it became one because the shape is deeply familiar and audiences follow it without effort. Applied to a brand, the discipline worth taking from it is who occupies which role — the customer is the one who changes, and the brand belongs somewhere alongside them.

Reach for it when

When a company’s story has the company at the centre of it and nobody outside is interested.

Where it stops

It is a description of myths, not a rule for them, and its universality has been criticised for decades. Followed literally it produces the same advertisement everyone else made.

Joseph Campbell, The Hero with a Thousand Faces, Pantheon, 1949.

A long-range goal paired with a vivid description of what it will be like to have arrived.

The second half is the part usually skipped and the part that does the work: not the target but a passionate, concrete description of the achieved state, written as if it were already true. Its function is translation — turning a figure nobody feels into a picture people can hold and repeat.

Reach for it when

When a vision statement has been written and nobody can describe what would be different if it happened.

Where it stops

Vividness is not evidence. A description compelling enough to move an organisation is exactly as easy to write for the wrong future as the right one.

James C. Collins & Jerry I. Porras, “Building Your Company’s Vision”, Harvard Business Review, 1996.

Start from the future you want, then work backwards to what would have had to happen for it to be reached.

It is the deliberate opposite of forecasting. Rather than extending today’s trends and accepting where they land, it fixes a desirable end state and asks what chain of conditions leads there. That reversal is what makes it a vision tool: it treats the future as something chosen rather than predicted.

Reach for it when

When planning keeps producing next year plus a percentage, and when a vision needs a path rather than only a picture.

Where it stops

It says nothing about feasibility. A perfectly reasoned path can lead back from a future that was never available.

John B. Robinson, “Energy Backcasting: A Proposed Method of Policy Analysis”, Energy Policy, 1982.

A mission holds when purpose, strategy, values and standards of behaviour all reinforce one another.

Four elements that must agree: why the company exists, the position it competes from, what it believes, and how people are actually expected to act. The diagnosis it enables is precise — a mission fails not because it is badly written but because one of the four contradicts the others, and it names which.

Reach for it when

When a mission statement is disbelieved inside the company and nobody can say exactly why.

Where it stops

Getting four elements to agree on paper is not the same as making them agree in the corridor. The model finds the contradiction; resolving it is management.

Andrew Campbell & Sally Yeung, “Creating a Sense of Mission”, Long Range Planning, 1991.

What is our mission, who is our customer, what does the customer value, what are our results, what is our plan.

Five questions in a deliberate order, each one difficult to answer honestly. The second and third carry the weight: naming the customer forces a choice, and answering what they value obliges you to admit you may not know. Everything else in the sequence depends on getting those two right.

Reach for it when

As the first conversation, before drafting anything. It works for a department as well as a company.

Where it stops

It is a discipline of questioning, not a method of answering. Everything depends on whether the room is willing to say something uncomfortable.

Peter F. Drucker, The Five Most Important Questions You Will Ever Ask About Your Organization, Jossey-Bass, 1993.

A checklist of the outside forces a business does not control: political, economic, social, technological, environmental, legal.

Its whole function is coverage. Left to themselves, teams scan the two categories they are already worried about, and the change that eventually matters arrives from the one nobody was watching. The letters have accumulated over the decades; the discipline of going through all of them has not changed.

Reach for it when

At the start of opportunity work, and once a year regardless of whether anything seems to be happening.

Where it stops

It produces a list, not a judgement. Without a second step that asks how likely and how consequential each item is, it is a wall of observations.

Developed from Francis J. Aguilar, Scanning the Business Environment, Macmillan, 1967.

Internal strengths and weaknesses set against external opportunities and threats.

Its lasting contribution is the axis nobody names: two of the boxes are about you and two are about the world, and an opportunity only counts if a strength can actually reach it. Used as four lists it is nearly worthless; used as a set of pairings — which strength serves which opportunity, which weakness is exposed to which threat — it still earns its place.

Reach for it when

To connect what a business has to what the market is doing, once both have been established separately.

Where it stops

It is the most abused tool in strategy. Filled in as four columns of adjectives, it generates agreement and no decisions.

Edmund P. Learned, C. Roland Christensen, Kenneth R. Andrews & William D. Guth, Business Policy: Text and Cases, Irwin, 1965; also attributed to Albert S. Humphrey at SRI.

The whole market, the part your model can serve, and the part you could realistically win.

Three nested figures, each narrower and each requiring a different argument. The discipline is in the narrowing: the total is usually available from a report, the serviceable portion requires you to state what your business actually cannot do, and the obtainable share requires you to name who else is there.

Reach for it when

Whenever an opportunity is being sized, and particularly when the number came from a market report.

Where it stops

Every layer is an estimate resting on the one above it. A top-down calculation can be made to produce almost any answer, and usually is.

Standard venture and corporate development practice; no single origin.

Products pass through introduction, growth, maturity and decline, and what to do about one depends entirely on where it is.

Spending to build awareness makes sense in the first stage and is waste in the third; defending share matters in maturity and is money lost in decline. The stage also predicts who else is in the market: growth attracts entrants, maturity produces price competition, and decline leaves whoever can run it cheapest.

Reach for it when

When an opportunity is being judged on its own merits without asking how old the market around it is.

Where it stops

The stages are obvious backwards and hard to identify at the time — a dip can be decline or a pause, and treating it as decline can make it one. Nothing in the model says how long a stage lasts.

Theodore Levitt, “Exploit the Product Life Cycle”, Harvard Business Review, 1965; the concept predates the article.

Profitability in an industry is set by five pressures, and rivals are only one of them.

The bargaining power of buyers and of suppliers, the threat of new entrants and of substitutes, and the intensity of rivalry. The reason it displaced simpler competitor analysis is its claim that the structure of an industry, not the skill of its participants, explains most of the difference in returns between industries — and that the force squeezing you is often not the company you think of as the competitor.

Reach for it when

When entering a market, when returns are falling across an industry rather than at one company, and whenever competitor analysis has become a list of rival firms.

Where it stops

It describes an industry at a moment and assumes reasonably clear boundaries. In markets where platforms make an industry a different shape every few years, the snapshot ages fast, and it has little to say about complementors or regulation.

Michael E. Porter, “How Competitive Forces Shape Strategy”, Harvard Business Review, 1979; Competitive Strategy, Free Press, 1980.

Competitors cluster into groups following similar strategies, and the real rivalry happens inside a group.

Plotting rivals on two dimensions that matter in the industry — breadth against price, say, or service level against reach — shows clusters rather than a scatter. Two findings follow: you compete hardest with your own cluster, and the empty spaces on the map are either an opportunity or a place where a business model does not work, and telling those apart is the actual work.

Reach for it when

When a competitor list has grown long and undifferentiated, and when looking for an unoccupied position.

Where it stops

Everything depends on choosing the two dimensions, and the choice is where the analysis is really done. A map on the wrong axes shows tidy clusters that mean nothing.

Michael S. Hunt, doctoral dissertation, Harvard, 1972; developed in Michael E. Porter, Competitive Strategy, Free Press, 1980.

Stop competing on the factors an industry already competes on, and change which factors are on the table at all.

Its working tool is a grid of four moves made together: which factors the industry takes for granted could be eliminated; which are over-served and could be reduced; which deserve to be raised well above the standard; and which the industry has never offered and could be created. The insistence on all four at once is the argument — cutting alone is cost-cutting, adding alone is feature creep, and only the combination changes what the offer is rather than how well it is done.

Reach for it when

When every competitor is improving the same handful of factors and the differences between them have stopped mattering to anyone buying.

Where it stops

The published cases are chosen after the fact, and there is no shortage of companies that removed the wrong factor and were never written up. It is a way of generating options, not evidence that any of them will work.

W. Chan Kim & Renée Mauborgne, Blue Ocean Strategy, Harvard Business School Press, 2005. Blue Ocean Strategy is a registered trademark of its owners and is named here only to refer to their work.

The competitor that displaces you usually starts by serving customers you were glad to lose.

A new entrant arrives with something cheaper and worse, takes the least profitable end of the market, and improves faster than the incumbents expect. The dilemma is that the incumbent’s response is rational at every step: serving its best customers and protecting its margins is exactly what good management says to do, right up until the entrant is good enough for everyone.

Reach for it when

When competitor analysis contains only firms of your own size, and when something cheap and unimpressive has appeared at the bottom of your market.

Where it stops

The term is used for any successful new company, which is not what it means, and the original case studies have been challenged. Most new entrants are simply competitors; disruption in the strict sense is rarer than the word suggests.

Joseph L. Bower & Clayton M. Christensen, “Disruptive Technologies: Catching the Wave”, Harvard Business Review, 1995; Christensen, The Innovator’s Dilemma, HBS Press, 1997.

Divide the market, choose deliberately which parts to serve, then decide what you will mean to them.

The sequence is the argument. Segmentation without targeting produces a description of everybody; targeting without segmentation is a guess; and positioning decided before either is a slogan looking for an audience. Its enduring value is that it forces an explicit act of exclusion in the middle step.

Reach for it when

Whenever the answer to who the customer is comes back as anyone who needs it.

Where it stops

It assumes segments are stable and reachable as groups. Where behaviour varies more within a segment than between segments, the whole structure describes nothing.

Wendell R. Smith, “Product Differentiation and Market Segmentation”, Journal of Marketing, 1956; developed by Philip Kotler, Marketing Management, from 1967.

A small number of specific, named, fictional people standing in for real patterns of behaviour.

Introduced in software design to stop teams building for an elastic user who conveniently wants whatever is easiest to build. The mechanism is specificity: a named person with a particular situation and a particular constraint cannot be quietly redefined mid-argument, and that is the entire benefit.

Reach for it when

When design or marketing decisions are being settled by whoever imagines the user most confidently.

Where it stops

Invented rather than researched, a persona is only a projection of the team, and demographic detail added for colour invites reasoning from stereotype.

Alan Cooper, The Inmates Are Running the Asylum, Sams, 1998.

Attention, interest, desire, action — the oldest description of the path from stranger to buyer.

More than a century old and still the underlying shape of most funnels drawn today. Its value now is diagnostic rather than descriptive: naming which step is failing is a far more useful question than asking why sales are down, and the four stages are enough structure to ask it.

Reach for it when

As a first cut at where a journey is losing people, before anything more elaborate is built.

Where it stops

It describes a linear path that almost nobody walks. Real journeys loop, pause for months and involve several people, and a funnel drawn this way makes that invisible.

E. St. Elmo Lewis, described in trade press from 1898; formalised in early advertising literature.

Buying is a loop, not a funnel: people add options as they go, and the loyalty loop after purchase is where the next sale is decided.

Based on a study of thousands of purchase decisions, it found the consideration set widening during evaluation rather than narrowing, and a substantial share of subsequent purchases bypassing consideration entirely. The consequence for spending is direct: the period after the sale earns more than the period before it, and receives a fraction of the budget.

Reach for it when

When a funnel model no longer matches what customers actually do, and when post-purchase experience has no owner.

Where it stops

It generalises across categories that behave very differently, and the shape it replaces the funnel with is still a simplification.

David Court, Dave Elzinga, Susan Mulder & Ole Jørgen Vetvik, “The Consumer Decision Journey”, McKinsey Quarterly, 2009.

Group audiences by how close they are to intent, and judge each stage by a measure that belongs to it.

Four audience states rather than four funnel steps, each with its own content and — the part that matters — its own measure of success. Its sharpest contribution is prohibitive: judging awareness activity by immediate conversion will kill it every time, and the framework exists largely to stop that.

Reach for it when

When early-stage marketing is being cut because it does not convert, and when one metric is being applied across the whole journey.

Where it stops

Built for digital measurement, and it assumes you can tell which state someone is in. Offline and long-cycle purchases fit it badly.

Avinash Kaushik, “See-Think-Do-Care: A Business Framework for Marketing”, Occam’s Razor, 2013.

The decisions a marketer controls, gathered into four headings — and three more once a service is involved.

Product, price, place and promotion, with people, process and physical evidence added for services in 1981. Its usefulness against channels is the reminder that a channel is a decision about place, and that place cannot be chosen independently of price and promotion without something breaking.

Reach for it when

When a channel decision is being made on its own, and when a promotion is planned that the pricing or the distribution cannot support.

Where it stops

It is organised around what the seller controls, which is why every attempt to reframe it from the buyer’s side has found followers. It is a checklist, not a strategy.

E. Jerome McCarthy, Basic Marketing, Irwin, 1960; Bernard H. Booms & Mary J. Bitner, 1981.

Statistical estimation of what each channel actually contributed, using aggregate data rather than tracking anyone.

Regression of outcomes on spend across channels, with adjustments for seasonality, price, distribution and the fact that advertising keeps working after it stops. It has returned to prominence for an unglamorous reason: it needs no individual-level tracking, which makes it robust to everything that has happened to cookies and consent.

Reach for it when

When channels are each claiming the same conversions, and when attribution based on tracking is no longer available or believable.

Where it stops

It needs years of data and real variation in spend to say anything, and correlation across channels that always move together cannot be separated however good the model is.

Developed in marketing econometrics from the 1960s; see Dominique M. Hanssens, Leonard J. Parsons & Randall L. Schultz, Market Response Models, 1990.

Content placed by what it is for — to entertain, inspire, educate or convince — and how close the reader is to buying.

Two axes: emotional against rational, and awareness against purchase. Its use is as an audit rather than a plan. Almost every content library, plotted onto it, turns out to be crowded into one quadrant, and the empty quadrants are usually the ones the audience needed.

Reach for it when

When auditing what already exists, before deciding what to make next.

Where it stops

It classifies formats and intents; it says nothing about quality or distribution. A balanced matrix of unread material is still unread.

Dan Bosomworth, The Content Marketing Matrix, Smart Insights, 2012.

Seventy per cent of effort on the dependable, twenty on what is promising, ten on what might not work at all.

A budgeting rule for a calendar rather than for a piece of content. Its function is protective: the ten per cent is the only part that produces anything new, and it is the first thing cut when a quarter looks difficult — so it is written into the plan as a proportion rather than left to survive on merit.

Reach for it when

When setting a year of publishing, and when experimental work keeps being postponed to the next quarter.

Where it stops

The proportions are a convention, not a finding. What they usefully enforce is that some deliberate share is set aside for risk, not that the share is seventy, twenty and ten.

Published as part of Coca-Cola’s “Content 2020” strategy, 2011; widely adopted since.

Three settings for how widely to sell: everywhere possible, through a chosen few, or through one.

Intensive distribution maximises availability and surrenders control over how the product is presented and priced. Exclusive distribution buys commitment and margin at the cost of reach. Selective sits between. The choice is not a channel preference but a strategic one, and it has to agree with the brand’s positioning or one of the two is lying.

Reach for it when

When adding a channel, and particularly when a marketplace or discounter is being added to a brand sold as premium.

Where it stops

It was formulated for physical goods and physical shelves. Online, availability is nearly free and the control problem takes a different form.

Standard in marketing channel theory; see Louis W. Stern & Adel I. El-Ansary, Marketing Channels, Prentice Hall, 1977.

In large sales, the questions that work are about situation, problem, implication and need-payoff — in that order.

Drawn from observation of thousands of sales calls rather than from theory. The finding that made it credible was that the closing techniques taught everywhere correlated with success in small sales and with failure in large ones, and that what distinguished successful large sales was the implication question: making the buyer state the cost of the problem themselves.

Reach for it when

Wherever the direct sales channel involves a considered purchase and more than one decision-maker.

Where it stops

The research is from the 1970s and 1980s, when the seller held the information. Buyers now arrive informed, and the situation questions that opened those calls now waste the meeting.

Neil Rackham, SPIN Selling, McGraw-Hill, 1988.

Map what the customer experiences and everything behind the counter that has to happen for it, on the same diagram.

A line of visibility separates what the customer sees from the work that supports it, with support processes below that again. Its diagnostic power comes from that line: most service failures are not failures of the visible step but of something underneath it that nobody had drawn.

Reach for it when

When a journey map records feelings and touchpoints but nothing about what produces them.

Where it stops

It is detailed and slow to produce, and it describes the service as designed. What is actually done in the branch on a Friday afternoon is frequently something else.

G. Lynn Shostack, “Designing Services That Deliver”, Harvard Business Review, 1984.

A company is judged in the brief encounters between a customer and a member of staff, and nowhere else.

Written from the turnaround of an airline, it argued that the impression of a company is created in millions of small interactions, none of which a manager attends. The management conclusion follows directly: authority has to sit with the person in the encounter, because the encounter is over long before anyone can be asked.

Reach for it when

When mapping a journey and deciding which points deserve real investment, and when frontline staff have responsibility without authority.

Where it stops

It comes from a high-contact service in the 1980s. Where most encounters are with software rather than people, the moments still exist but the conclusion about authority has to be rebuilt.

Jan Carlzon, Moments of Truth, Ballinger, 1987.

One question — how likely are you to recommend us — scored so that the indifferent count for nothing.

Its real contribution was operational rather than statistical: one question gets answered, gets asked continuously, and produces a number an organisation can be organised around. The follow-up question asking why is where the usable information actually is, and it is the part most often dropped.

Reach for it when

When feedback is collected in long questionnaires that few complete and nobody reads.

Where it stops

The claimed link to growth has not survived independent replication, and the scoring throws away information. Treated as a trend on the same population it is useful; treated as a benchmark across companies or countries it is not.

Frederick F. Reichheld, “The One Number You Need to Grow”, Harvard Business Review, 2003. Net Promoter and NPS are registered trademarks of their owners and are named here only to refer to their work.

Ask how hard the customer had to work, because reducing effort predicts loyalty better than delighting does.

The underlying study found that exceeding expectations in service barely moved loyalty, while making things difficult destroyed it — and that most service interactions are a chance to lose a customer rather than to win one. Measuring effort points attention at the same place: removing obstacles rather than adding delight.

Reach for it when

In support and self-service, where the ambition of delighting people is expensive and mostly wasted.

Where it stops

It measures one interaction, not the relationship. A product people love can be effortful, and effort is not always something the customer wants removed.

Matthew Dixon, Karen Freeman & Nicholas Toman, “Stop Trying to Delight Your Customers”, Harvard Business Review, 2010.

Poor service is the gap between what customers expected and what they perceived, and that gap has four causes behind it.

Not knowing what customers expect; specifying a standard that does not match it; not delivering to the standard; and promising something different in communication. Locating which of the four is open changes what you do — three of them are internal problems and one of them is a marketing problem, and they are routinely confused.

Reach for it when

When service scores are poor and the response is more training for the people at the counter.

Where it stops

The associated questionnaire and its five dimensions have been criticised for decades, particularly the expectations half. The gap structure has aged better than the instrument.

A. Parasuraman, Valarie A. Zeithaml & Leonard L. Berry, “A Conceptual Model of Service Quality”, Journal of Marketing, 1985.

A failure handled well can leave a customer more committed than if nothing had gone wrong — but only sometimes.

The paradox is real and conditional: it holds for first failures, when recovery is fast and generous, and it does not survive repetition. What is robust in the literature underneath it is less dramatic and more useful — speed matters more than compensation, and being told what happened matters more than either.

Reach for it when

When designing what happens after something goes wrong, and when deciding how much authority the first person to hear about it should have.

Where it stops

It is not a licence to fail deliberately. The effect is unreliable, the second failure removes it entirely, and prevention is cheaper than any recovery.

Michael A. McCollough & Sundar G. Bharadwaj, “The Recovery Paradox”, AMA Educators’ Conference, 1992; extensive literature since.

Ten general principles for interface design, used as a checklist to find problems without recruiting users.

Visibility of system status, match with the real world, user control, consistency, error prevention, recognition over recall, and the rest. They were derived by factor analysis of hundreds of real usability problems, which is why they are broad — each one covers a family of failures rather than a specific rule.

Reach for it when

As a first inspection pass before user testing, where a few evaluators will find most of the serious problems cheaply.

Where it stops

Heuristic evaluation finds violations of principles, not the things real users actually stumble over. It reduces the need for testing; it does not remove it.

Jakob Nielsen, “Enhancing the Explanatory Power of Usability Heuristics”, CHI ’94; refined 1994–2020.

Two cycles of widening then narrowing: first to find the right problem, then to find the right answer to it.

Discover, define, develop, deliver. The reason there are two diamonds rather than one is the whole argument: teams converge on a solution before they have converged on a problem, and the first diamond exists to make that difficult. The widening halves are the ones under pressure to be skipped.

Reach for it when

At the start of any redesign, particularly one that began with a proposed solution.

Where it stops

It describes a shape, not a practice. Drawn on a plan while the team works the way it always did, it changes nothing.

UK Design Council, 2005; revised as the Framework for Innovation, 2019.

A behaviour happens when motivation, ability and a prompt arrive at the same moment — and fails if any is missing.

Motivation and ability trade off against each other: something easy enough needs little motivation, something hard needs a great deal. The design consequence is the useful one, and it runs against instinct — when people do not do something, making it easier works more reliably than trying to make them want it more.

Reach for it when

When a step in a flow is being abandoned and the proposed fix is more persuasive copy.

Where it stops

It is a model of the moment, not of habit or of whether the behaviour was good for the person. Its close relation to persuasive design makes the ethical question unavoidable rather than optional.

B. J. Fogg, “A Behavior Model for Persuasive Design”, Persuasive ’09, Stanford, 2009.

Sort customers by how recently they bought, how often, and for how much — three fields that predict a great deal.

It long predates machine learning and remains an unreasonably strong baseline. Recency does most of the work: how long ago someone last bought predicts whether they will buy again better than almost anything else on file, and all three fields are already in every order table.

Reach for it when

As the first personalisation anyone builds, and as the baseline any more sophisticated model must beat.

Where it stops

It describes what someone did, not what they want. It cannot help with a new customer, and it will keep recommending more of what they already bought.

Long-standing direct marketing practice; formalised in Jan Roelf Bult & Tom Wansbeek, “Optimal Selection for Direct Mail”, Marketing Science, 1995.

Recommend to someone what people who behaved like them went on to choose, without needing to know anything about the item.

The insight is that similarity of behaviour carries more information than description of content, and it needs no understanding of what is being recommended at all. Two problems have followed it ever since: it has nothing to say about a new user or a new item, and it narrows what people are shown to a reflection of what they already did.

Reach for it when

Where there is enough behavioural data for patterns to be real, and as the standard against which content-based approaches are judged.

Where it stops

It amplifies the popular and the already-chosen. Left alone it will recommend a company’s catalogue into a small corner of itself.

David Goldberg, David Nichols, Brian M. Oki & Douglas Terry, “Using Collaborative Filtering to Weave an Information Tapestry”, Communications of the ACM, 1992.

Buying repeatedly and being committed are different things, and only one of them survives a better offer.

Repeat purchase can come from habit, convenience or lack of an alternative, and looks identical in the data to genuine commitment. The distinction matters most where it is inconvenient: a loyalty programme that rewards repeat purchase may be paying people who were never going anywhere, while doing nothing about the ones who were.

Reach for it when

Before designing or renewing a loyalty scheme, and when repeat-purchase figures are being read as evidence of loyalty.

Where it stops

Attitudinal loyalty can only be measured by asking, with all the softness that implies. The distinction is easy to state and difficult to operationalise.

Jacob Jacoby & Robert W. Chestnut, Brand Loyalty: Measurement and Management, Wiley, 1978.

Internal quality produces satisfied employees, who produce satisfied customers, who produce profit — in that order.

A chain of proposed links, each supported by data from service businesses, ending in the claim that customer loyalty is worth far more than customer satisfaction alone and that employee retention is where it starts. For loyalty work it relocates the problem: the strongest lever on whether customers stay may be several steps upstream of anything a programme can offer.

Reach for it when

When loyalty is being pursued entirely through incentives and the service itself is unchanged.

Where it stops

The links are correlational, drawn largely from high-contact service businesses. The direction of causation between employee and customer satisfaction is not as settled as the diagram implies.

James L. Heskett, Thomas O. Jones, Gary W. Loveman, W. Earl Sasser & Leonard A. Schlesinger, “Putting the Service-Profit Chain to Work”, Harvard Business Review, 1994.

People deliberately research in one channel and buy in another, and the pattern is stable enough to plan around.

The study measured the movement rather than assuming it, and found it driven by what each channel is genuinely better at — one for comparison and detail, another for reassurance or immediacy. It gives omnichannel work a concrete goal: not to make every channel do everything, but to let a customer cross between them without losing their place.

Reach for it when

When channels are measured separately and each is being judged on the sales it closes.

Where it stops

It was measured before smartphones changed the mechanics of it, and channel attribution remains genuinely unsolved. It names the behaviour better than it quantifies it today.

Peter C. Verhoef, Scott A. Neslin & Björn Vroomen, “Multichannel Customer Management: Understanding the Research-Shopper Phenomenon”, International Journal of Research in Marketing, 2007.

Measure across four perspectives at once, so financial results are read alongside what produces them.

Financial, customer, internal process, and learning and growth. The design argument is causal rather than decorative: the last three are the leading indicators of the first, and a company reading only financial results is steering by an outcome that reports on decisions taken a year ago.

Reach for it when

When a measurement set is entirely financial, or when a KPI list has grown without anyone asking what causes what.

Where it stops

It is heavy, and organisations tend to keep the four boxes and drop the causal links between them — at which point it is a longer report rather than a better one.

Robert S. Kaplan & David P. Norton, “The Balanced Scorecard — Measures That Drive Performance”, Harvard Business Review, 1992.

When a measure becomes a target, it stops being a good measure.

Originally an observation about monetary policy, generalised since to any managed indicator. The mechanism is not dishonesty: a measure is a proxy for something you care about, and optimising the proxy will eventually pull it away from the thing it stood for, whether or not anyone intends it.

Reach for it when

Before attaching consequences to any indicator, and whenever a metric has improved dramatically while the outcome behind it has not.

Where it stops

It is a warning, not a method. It does not say which measures will corrupt or how to prevent it — only that they will, and that unmeasured judgement remains necessary.

Charles A. E. Goodhart, 1975; this wording from Marilyn Strathern, “Improving Ratings”, European Review, 1997.

Quality is not one property but several — completeness, uniqueness, timeliness, validity, accuracy, consistency.

Splitting quality into named dimensions makes it arguable and measurable. It also settles a common confusion: data can be entirely valid and entirely wrong, since validity asks whether a value is of the right form and accuracy asks whether it describes reality, and only one of the two is cheap to check.

Reach for it when

When setting up collection, and when a dataset is described as bad without anyone saying in what way.

Where it stops

Measuring the dimensions costs real effort, and perfection in all six is neither affordable nor necessary. Which dimensions matter depends on what the data is for.

DAMA UK Working Group, The Six Primary Dimensions for Data Quality Assessment, 2013.

Every source of error in a survey, gathered into one frame so they can be traded off against each other.

Coverage, sampling, non-response and measurement error all reduce accuracy, and they compete for the same budget. The insight that changed practice is that non-response bias, not sample size, usually dominates — which means a larger survey of the same self-selecting people buys precision around a wrong number.

Reach for it when

Before commissioning any survey, and when reading someone else’s.

Where it stops

Several of the errors cannot be measured from inside the survey itself. It tells you what to worry about rather than how much.

Robert M. Groves, Survey Errors and Survey Costs, Wiley, 1989.

Randomised assignment is the only method that separates what your change caused from what would have happened anyway.

The logic is a century old and the application is recent. What the web experimentation literature added is a catalogue of the ways it goes wrong in practice — peeking at results, running many tests until one passes, and the consistent finding that the large majority of confidently expected improvements turn out to be nothing.

Reach for it when

Whenever a change can be given to some people and not others, and before believing any before-and-after comparison.

Where it stops

It answers narrow questions about small changes over short horizons. Long-term effects, changes that must ship to everyone, and anything with few users are outside it.

Ronald A. Fisher, The Design of Experiments, Oliver & Boyd, 1935; Ron Kohavi, Diane Tang & Ya Xu, Trustworthy Online Controlled Experiments, Cambridge University Press, 2020.

Group people by when they arrived, then follow each group separately instead of averaging them together.

An aggregate figure mixes people who have been customers for three years with people who arrived last week, and the mixture hides the trend. Following each intake on its own shows whether what you changed made later arrivals behave differently — which is the question, and one that no total can answer.

Reach for it when

Whenever retention, repeat purchase or usage over time is being judged from a single overall number.

Where it stops

Cohorts differ for reasons other than what you changed — the season they arrived in, the campaign that brought them. It shows a difference without explaining it.

Standard demographic method; brought into business analytics through subscription and web practice.

Work backwards from an observed problem through the categories of cause that could produce it.

Branches for the standard families — people, method, machine, material, measurement, environment — each expanded until the causes are specific enough to test. Its value is coverage under pressure: a group looking for a cause converges on the first plausible one within minutes, and the categories force the other five to be considered.

Reach for it when

When performance has moved and the explanation was agreed on before anyone looked.

Where it stops

It generates candidate causes; it does not weigh them. Every branch is a hypothesis and the diagram gives no way to tell which one is true.

Kaoru Ishikawa, Guide to Quality Control, JUSE, 1968.

State the answer first, then the reasons that support it, then the evidence underneath each reason.

Developed at McKinsey for written recommendations. Ideas are grouped so that each level summarises the one below, and the whole thing is read top-down. It works because a reader who has the conclusion can evaluate the argument as it arrives, and a reader who does not is merely waiting.

Reach for it when

For any report or recommendation, and especially for the ones that build carefully to a conclusion on the last page.

Where it stops

It structures a case you have already reached. It is a way of communicating a conclusion, not of finding one, and used too early it hardens a view before the analysis is done.

Barbara Minto, The Pyramid Principle, Pitman, 1973.

Most of the ink in a graphic should be showing data, and the rest is usually in the way.

Part of a broader argument that graphical excellence is telling the truth about numbers with the least distortion. Its practical instructions are unusually concrete — remove the decoration, remove the redundancy, and above all do not let a chart imply a relationship the data does not contain.

Reach for it when

On any recurring report, where a small improvement is paid back every week.

Where it stops

Taken to its conclusion it produces charts that are austere to the point of being hard to read, and later research suggests some decoration aids memory. It is a corrective, not a rule.

Edward R. Tufte, The Visual Display of Quantitative Information, Graphics Press, 1983.

A complete map of the disciplines data management consists of, and how governance sits at the centre of them.

Eleven knowledge areas — architecture, modelling, storage, security, integration, quality and the rest — arranged around governance. Its use to an organisation starting out is mostly as a map of what exists: it prevents the common error of building a governance policy that addresses two of the eleven and calls the job done.

Reach for it when

When establishing governance, and as a checklist of what has not yet been considered.

Where it stops

It is a reference work of considerable weight, written for large organisations. Adopted wholesale by a small one it produces a great deal of ceremony.

DAMA International, DAMA-DMBOK: Data Management Body of Knowledge, Technics Publications, 2009; second edition 2017.

Treat data as a product owned by the domain that produces it, with governance agreed centrally and applied locally.

A reaction to the central data team that becomes a queue. Ownership moves to the domains that understand the data, each publishing it as a product with defined quality and an interface, while standards are set once and enforced everywhere. Governance becomes federated rather than either centralised or absent.

Reach for it when

When a central data function is the bottleneck, and when nobody can say who is responsible for a given dataset being right.

Where it stops

It presupposes domains mature enough to own a product, and it is routinely adopted as a technology purchase rather than the organisational change it actually is.

Zhamak Dehghani, “How to Move Beyond a Monolithic Data Lake to a Distributed Data Mesh”, martinfowler.com, 2019.

The longest chain of dependent tasks sets the finish date, and only that chain does.

Everything else has slack. The practical consequences are unintuitive and reliably ignored: adding people to a task off the critical path changes nothing, a delay on the path is a delay to the project, and the path itself moves as work progresses, so it has to be recalculated rather than drawn once.

Reach for it when

On any project where tasks genuinely depend on one another and a date has been promised.

Where it stops

It assumes known tasks and known durations. Estimates carry padding at every step, and the method has nothing to say about work whose shape is discovered while doing it.

James E. Kelley & Morgan R. Walker, “Critical-Path Planning and Scheduling”, Eastern Joint Computer Conference, 1959.

Work proceeds in stages separated by decision points where a project can be stopped.

Each gate asks the same three questions — is it still worth doing, are we doing it well, and what would justify continuing — with criteria agreed before the answer is known. The gate that matters is the one where a project is killed, and its absence is why organisations carry projects nobody believes in for years.

Reach for it when

Where several initiatives compete for the same money and none of them ever ends.

Where it stops

Applied rigidly it is slow and rewards documentation over evidence. Cooper’s own later work loosens the gates considerably for exactly that reason.

Robert G. Cooper, Winning at New Products, Addison-Wesley, 1986.

Fixed short cycles producing something usable, with the plan reconsidered at the end of each one.

A deliberately small set of roles, events and artefacts built on the assumption that requirements will be wrong and are best corrected by evidence rather than by analysis. Its most valuable and least practised element is the retrospective, which is the only part that changes how the team works rather than what it produces.

Reach for it when

Where what should be built is genuinely uncertain and can be delivered in pieces.

Where it stops

It is a container, not a method — it says nothing about how to build anything well. Adopted as ceremonies without the authority to change direction, it becomes weekly status reporting with new vocabulary.

Ken Schwaber, “SCRUM Development Process”, OOPSLA ’95; Schwaber & Sutherland, The Scrum Guide, from 2010.

For each task, name who does it, who answers for it, who must be asked and who must be told.

The distinction that earns its keep is between responsible and accountable: several people can do the work, but exactly one answers for whether it happened. Most of the value appears while filling it in, when it turns out two people believed they were accountable for the same thing, or nobody was.

Reach for it when

On any piece of work involving more than one team, and immediately after something fell between two of them.

Where it stops

It allocates existing work; it does not question whether the work is right. Kept as a document rather than an agreement it becomes something to point at after the fact.

Long-standing practice in project management; documented in the PMBOK Guide, Project Management Institute.

Every task drawn as a bar against a calendar, so that duration, overlap and sequence can be seen at once.

More than a century old and still the default way work is shown over time. What it does well is expose two things that are invisible in a task list: how much is happening at the same moment, and which things cannot start until something else finishes. Adding the dependencies is what turns it from a picture into a plan.

Reach for it when

When a schedule exists only as a list of dates and nobody can see where the load falls.

Where it stops

It shows the plan, not the risk in it, and it invites false precision — a bar drawn to the day looks like a commitment even when the estimate behind it was a guess. For work that is discovered as it goes, it ages within a week.

Developed by Henry L. Gantt around 1910–1915; a similar chart was published by Karol Adamiecki in 1896.

New things spread through a population in a predictable order, from a small group of the venturesome to a reluctant last.

Synthesised from hundreds of diffusion studies across agriculture, medicine and technology. Alongside the adopter categories it names the properties that make something spread quickly — relative advantage, compatibility, simplicity, whether it can be tried cheaply, and whether adoption is visible to others — which is the part most useful for timing a launch.

Reach for it when

When deciding when to enter, and when early enthusiasm is being read as evidence of a mainstream market.

Where it stops

The curve is drawn from adoptions that succeeded. It is a description of how spreading works, not a prediction that a particular thing will spread.

Everett M. Rogers, Diffusion of Innovations, Free Press, 1962.

Between the enthusiasts and the mainstream there is a gap, because the two groups buy for incompatible reasons.

Early adopters buy a change and tolerate rough edges; the early majority buy a solved problem and take their evidence from people like themselves — who, by definition, have not bought yet. Moore’s recommendation is to attack one narrow segment completely enough to become its obvious choice, and only then widen.

Reach for it when

When a product has passionate early users and growth has stalled without anyone being able to say why.

Where it stops

Drawn from business technology in a particular era, and the segment-at-a-time prescription is easier to state than to execute. Not every category has a chasm.

Geoffrey A. Moore, Crossing the Chasm, HarperBusiness, 1991.

Observe, orient, decide, act — and the advantage goes to whoever completes the cycle faster.

Developed from air combat and generalised to competition. Orientation is the part that carries the weight and the part usually dropped: it is where existing beliefs distort what was observed, and Boyd’s argument was that most failures are failures of orientation rather than of decision.

Reach for it when

When an opportunity has a window on it, and when the organisation’s decision cycle is visibly slower than events.

Where it stops

Speed applied to a wrong orientation only reaches the wrong place sooner. It is frequently quoted as an argument for acting fast, which is close to the opposite of what it says.

John R. Boyd, “Destruction and Creation”, 1976, and the Patterns of Conflict briefings, 1977–1986.

A small investment that buys the right to act later, without the obligation, has value that ordinary appraisal misses.

Applying option reasoning to real investments captures something discounted cash flow cannot: under uncertainty, the ability to wait, to expand or to abandon is itself worth money. It reframes a pilot project as the purchase of an option rather than a small version of a commitment.

Reach for it when

When an opportunity is uncertain and expensive, and the choice is being framed as now or never.

Where it stops

The formal valuation needs inputs that rarely exist outside financial markets. Used as a way of thinking it is valuable; used as a calculation it produces false precision.

Stewart C. Myers, “Determinants of Corporate Borrowing”, Journal of Financial Economics, 1977; developed in Avinash Dixit & Robert Pindyck, Investment Under Uncertainty, 1994.

Several internally consistent futures, written out in detail, instead of one forecast with a margin of error.

The purpose is not to predict but to make managers capable of recognising a future when it starts arriving. Wack’s account is explicit that the scenarios matter less than the change in what executives are prepared to see — which is why scenarios written by consultants and read once accomplish nothing.

Reach for it when

When planning rests on a single projection, and where the uncertainty is structural rather than statistical.

Where it stops

It is slow and demanding, and its value is destroyed by treating the scenarios as a menu with one marked most likely.

Pierre Wack, “Scenarios: Uncharted Waters Ahead” and “Shooting the Rapids”, Harvard Business Review, 1985.

Assume the plan has already failed a year from now, and have everyone write down why.

The reversal does the work. Asking what could go wrong invites loyal reassurance; stating that it did go wrong and asking for the reasons gives people permission to say what they already suspected. In the research behind it, prospective hindsight of this kind substantially increased the number of causes people could identify.

Reach for it when

Immediately after a plan is agreed and before it is committed to — it takes under an hour.

Where it stops

It surfaces concerns; it does not weigh them or fix them. Run as a ritual with no obligation to change the plan, it is worse than not running it.

Gary Klein, “Performing a Project Premortem”, Harvard Business Review, 2007; building on Mitchell, Russo & Pennington, 1989.

Replace the fixed annual budget with rolling forecasts and targets set against actual conditions.

The case against the annual cycle is that it fixes a plan to a year that has not happened, negotiates resources once, and then rewards hitting a number agreed before anyone knew anything. What replaces it is a rhythm: forecasts refreshed continuously, resources released when needed, and performance judged relative to how the market actually turned out.

Reach for it when

When the annual plan is out of date by March and everyone spends the rest of the year explaining variance against it.

Where it stops

It requires giving up a great deal of central control, which is why most adoptions keep the budget and add the forecasting on top — the worst of both.

Jeremy Hope & Robin Fraser, Beyond Budgeting, Harvard Business School Press, 2003.

One monthly cycle in which sales, operations and finance agree a single set of numbers for the months ahead.

Demand is reviewed, supply is reviewed, the gap between them is resolved, and leadership signs off on one plan rather than the three that departments would otherwise run on. The discipline is the cadence: the same meeting, the same horizon, every month, so disagreements are settled on a schedule instead of during a crisis.

Reach for it when

As the backbone of a year wheel, and wherever sales forecasts and production plans are visibly different documents.

Where it stops

It needs honest forecasts, and the meeting is only as good as the willingness to say a number is wrong. Where it becomes a reporting ritual, it consumes a day a month and settles nothing.

Developed from Oliver Wight’s work on manufacturing planning in the 1980s; now standard practice in operations management.