Marketing
in a Hurry

The Whole Thing in One Page

Marketing has a visibility problem. What most people can see is the communication: a campaign, a sponsored result, a launch film, a discount email. The decisive work often happened months earlier, when somebody chose which customers mattered, decided what the product would mean beside its alternatives, set the terms of the bargain and selected the route by which the customer could buy it.

That broader view is not an attempt to claim the whole company for marketing. It is a way of locating the discipline correctly. Marketing is the design and management of a market-facing system. It connects what an organisation can do with what a group of buyers will choose, pay for, receive and choose again. Advertising can bring attention to that system. Branding can make it recognisable and meaningful. Sales can convert a prospect through direct human contact. Marketing must make those specialist activities point at a coherent exchange.

The subtitle gives the three most visible levers. Positioning determines the comparison. An offer turns a product or service into a complete bargain, including price, proof, risk, terms and service. Channels determine where the bargain can be encountered and completed. But those choices only work after a harder one: what market are you trying to win, and which customers inside it are worth designing around?

Markets are not invented from nothing. People already have needs, habits, budgets, technologies and alternatives. Yet markets are not fixed containers either. Firms create categories, bundle features, set price fences, subsidise one side of a platform, put products beside particular competitors and decide what becomes easy to compare. Marketing responds to demand while also helping organise the choice through which demand appears.

That is why the product and the promotion cannot be treated as separate from the rest of the bargain. Xerox did not make the 914 copier easier to adopt merely by describing it better. Leasing and usage-linked charging changed the risk and commitment. Rolls-Royce's Power-by-the-Hour support model changed what an aircraft operator was buying from an engine maker. Supermarkets, app stores, wholesalers, search engines and marketplaces do something equally consequential: they are not pipes. They rank, recommend, bundle, tax and sometimes own the customer relationship.

The loop continues after purchase. Acquisition can destroy value if the wrong customers churn. Retention can destroy value if the firm keeps unprofitable customers through discounts or obstruction. A healthy marketing system therefore tracks cohorts, contribution, experience and repeat behaviour, not merely clicks and conversions.

Measurement is the final trap. A dashboard can tell you what happened inside its field of view without telling you what your activity caused. Attribution allocates observed credit. Causal measurement asks what would have happened otherwise. The difference matters because an organisation that optimises whatever is easiest to count will gradually redesign its market around the metric.

The central idea is therefore not a funnel. It is a set of linked choices under competition: define the market, establish the comparison, construct the offer, choose the price and channel, create and capture demand, deliver the experience, then learn causally enough to change the right thing.

That is the book.

Why You Should Care

A business can waste a surprising amount of money proving that people will click on something they should never have been offered.

That sounds like a problem in advertising. Often it is a problem that existed before the advert. The wrong customer was targeted because the segment was easy to identify. The product was compared with the wrong alternative. The price was inherited from a cost spreadsheet. The free trial attracted users with no reason to stay. The marketplace delivered volume while taking the margin and the customer data. Marketing activity can be busy, measurable and commercially wrong at the same time.

The Xerox 914 shows why this subject begins before communication. Introduced in 1959, it was a breakthrough plain-paper copier, but it was also large, unfamiliar and expensive. Xerox used leasing and copy-based charging so that offices could gain access without making the same large purchase commitment up front. The machine mattered. So did the architecture around the machine. A different bargain changed the adoption problem.

The same principle appears everywhere. A low-cost airline is not a conventional airline with cheaper adverts. Its route network, airport choices, turnaround, service level, ancillary pricing and booking system create the proposition together. A software product sold directly to a five-person team faces a different marketing problem from the same software entering an enterprise through procurement, security review and implementation partners. A grocery brand moved from a supermarket shelf to a subscription service changes who controls assortment, comparison, data, replenishment and cancellation.

This matters to customers because marketing changes the environment in which choices are made. A category label tells you which alternatives belong together. A default decides what happens when you do nothing. A bundle hides some comparisons and reveals others. A free trial shifts initial risk towards the seller, while automatic renewal shifts inertia towards the seller later. Rankings and recommendations can make popularity partly self-reinforcing. None of this means buyers are puppets. It means choices have architecture.

It matters to managers because the usual departmental boundaries conceal causal relationships. Product teams can add features that reduce adoption. Finance can raise a headline price while improving a more important term. Operations can create a delivery promise that positioning cannot support. Sales can close customers whose implementation costs make them unprofitable. Marketing is useful when it joins these consequences rather than merely supplying leads to the next department.

It also matters because growth is an ambiguous word. A campaign can increase orders while lowering contribution. A discount can improve this month's conversion while teaching customers to wait next month. A marketplace can accelerate acquisition while making the business dependent on a gatekeeper that can change fees or ranking rules. Retention can rise because customers are delighted or because exit has been made irritating. The same metric can describe different mechanisms.

The discipline is therefore less about finding a clever message than about making better commercial hypotheses. Which customer and situation are we serving? What are they doing instead? Why should our difference matter? Which parts of the offer reduce risk or effort? What does the channel require in return for reach? What would happen without this activity? What happens to the customer after the first purchase?

There is no universal answer. Consumer staples, industrial equipment, luxury goods, marketplaces, professional services and subscription software have different buying processes and economics. Marketing is valuable precisely because it provides a way to reason across those differences without pretending they are the same.

Once you see the subject this way, promotion shrinks to its proper size. Attention is useful when the underlying exchange deserves more of it. The rest of the book is about building that exchange and learning whether it works.

The Core Ideas

1. A Market Is Defined Around a Decision

Markets look objective from a distance. There is a smartphone market, a mortgage market, a market for accountancy software. Get closer and the borders move.

A family choosing how to travel from London to Edinburgh can compare a flight, a train, a car and, in some circumstances, not travelling at all. An airline analysing that trip competes with more than other airlines. Yet the same airline negotiating a global corporate travel contract may compete inside a procurement category where rail is almost irrelevant. The physical service is similar. The decision, alternatives, buyer and criteria are different.

That is why marketing begins by defining the decision rather than collecting a demographic. A useful market description names a customer, a situation, a problem or desired progress, the alternatives and the constraints on action. It also distinguishes the user from the buyer, payer and veto holder where those roles differ. In business markets, a person who loves the product may still be unable to buy it. In household markets, the purchaser may be buying for somebody else.

Segmentation is the next act of judgement. A segment is useful when a group responds differently to something the organisation can change. Age, income and postcode can be useful locating variables, but they are not automatically explanations. The important difference may be urgency, desired outcome, risk tolerance, expertise, service requirement, willingness to switch or preferred route to purchase. If two groups look different in a database but should receive the same proposition, the segmentation has done little work.

Targeting then creates the part managers often resist: exclusion. A service optimised for speed will disappoint customers who need extensive customisation. A product built for experts may be hostile to beginners. An enterprise-grade purchasing process may make no sense for a small team. Trying to accommodate every possible buyer can make the proposition vague, the operations expensive and the evidence hard to interpret.

The word choice needs care. Firms do not choose markets with unlimited freedom. They inherit capabilities, regulation, competitors, technology, distribution and history. Demand can be latent, seasonal, habitual or constrained by income. A company cannot declare a profitable segment into existence. What it can do is decide which pattern of demand it will investigate, which problems it will solve and which trade-offs it will accept.

Research helps when it is designed to disconfirm as well as confirm. Interviews reveal vocabulary, anxieties and decision rules people can articulate. Observation reveals workarounds and friction they may not mention. Search behaviour, transaction data, lost deals, service complaints and sales calls expose other parts of the problem. Stated purchase intentions are informative but imperfect predictors of later behaviour, especially when real money, time and competing priorities enter.

A serious research programme therefore asks several kinds of question. What happened? Why did the buyer say it happened? What alternative was present? What did the buyer sacrifice? What changed across segments or occasions? What behaviour would make our current theory false? The last question protects research from becoming theatre in which customers are invited to praise an idea already funded.

There is a second complication. Marketing does not merely discover demand. Categories and comparisons help shape it. Before sports drinks became a familiar category, buyers did not enter shops with a stable, pre-existing concept of one. Before cloud software, many business buyers thought in terms of licences and installed systems. Firms can teach people to notice a problem, name a category or compare on a new dimension. They still operate against real needs and budgets, but the market's language and boundaries are partly made through competition.

So the first marketing decision is not “Who might buy this?” It is “Whose decision are we trying to win, against what, and on what evidence?” Every later choice depends on the answer.

This also changes how market size should be treated. Top-down reports can show that a category is large, but they can conceal whether the organisation can reach the relevant buyers, whether those buyers switch often enough to matter and whether the problem is severe enough to justify action. A smaller market with frequent, costly pain can be more attractive than a huge category in which customers are satisfied with inertia. Market size is therefore not merely population multiplied by price. It is the amount of reachable, winnable demand under a specific proposition and competitive situation.

2. Positioning Determines the Comparison

A product does not enter an empty mind. It enters a comparison.

Positioning is the disciplined choice of which comparison should organise the buyer's understanding and why the offer deserves preference inside it. This is deeper than a tagline. A tagline can express a position. It cannot rescue a position that the product, price and channel contradict.

The first task is to identify the alternatives customers already use. These can be direct competitors, internal workarounds, doing nothing, hiring a person, building something in-house or spending the budget elsewhere. If a team buys software to reduce manual reporting, a spreadsheet may be the meaningful competitor even if no software vendor lists Microsoft Excel on its battlecard. The customer decides the comparison set, not the seller's industry taxonomy.

Next comes category or frame. The buyer needs to know what kind of thing this is before most differences become interpretable. A new offer that resembles a familiar category benefits from the expectations attached to it, but also inherits the category's minimum requirements. A hotel can differentiate through communal spaces, design or local programming, yet the guest still expects a secure room and a functioning booking process. Distinction matters after basic competence is credible.

Then come the differences that matter. A strong point of difference is valued by the target customer, supported by evidence and connected to capabilities the organisation can sustain. The most defensible differences often arise from systems rather than isolated features. Southwest Airlines became a canonical strategy example because its short-haul model involved linked choices about aircraft, routes, turnaround, service and cost. IKEA's flat packs, warehouse stores, self-service journey, product design and customer assembly form another linked system. A rival can imitate one visible feature more easily than the whole activity set.

Trade-offs give positioning credibility. If a company says it is best for every customer, every use case and every service level, the claim contains no reason to believe it. A position becomes legible when it states where the offer is strongest and what follows from that choice. “Fast setup for small teams” is informative partly because it implies that a complex global implementation is not the design centre.

Positioning also affects what gets built. If the position rests on being the lowest-friction option, onboarding, support, packaging and product complexity must be judged against that promise. If it rests on high assurance, the evidence, service and channel must support confidence. Marketing fails when positioning is treated as copy written after product decisions rather than a hypothesis that coordinates them.

There is an ethical dimension too. Positioning selects what becomes salient. A lender can foreground a monthly payment and make total cost less prominent. A food product can choose a category frame that makes one nutritional comparison easy and another obscure. A platform can label a sponsored placement in a way that blurs the distinction between relevance and payment. Framing is unavoidable, which is precisely why transparency matters.

Positions decay. Competitors copy. Customer expectations rise. A differentiator can become a point of parity, and a category can split into subcategories with different criteria. The response is not necessarily rebranding. It may require a new offer, new capabilities or the admission that the old segment is no longer attractive.

A position, then, is a claim about a live choice. It answers: compared with what, for whom, on which dimension, with what proof and with which trade-offs?

The practical test is behavioural. Does the position help the right customer recognise the offer quickly, understand why it differs and predict the experience they will receive? If the market can repeat the slogan but still cannot tell when to choose the product, the communication may be memorable while the position remains weak. Distinctive language is useful only when it compresses a real commercial difference.

3. The Offer Is the Complete Bargain

The product is what the organisation makes. The offer is what the customer must say yes to.

That difference explains why excellent products can be difficult to buy and ordinary products can be commercially powerful. The offer includes the product or service, but also price, packaging, financing, terms, delivery, implementation, support, guarantees, proof, switching effort and the allocation of risk. These elements determine whether the customer can receive the promised value without an unacceptable sacrifice.

Rolls-Royce's Power-by-the-Hour arrangement made the distinction unusually visible. Beginning in 1962 for the Viper business-jet engine, support could be charged on a fixed cost per flying hour with maintenance responsibility bundled into the arrangement. The customer was buying more predictable availability and shifting part of the maintenance risk. The engine did not disappear. The economic meaning of owning and operating it changed.

Xerox's 914 created a related adoption problem. The plain-paper copier was valuable but expensive and unfamiliar. Leasing and usage-linked charges reduced the up-front commitment and allowed use to reveal value over time. The innovation in the bargain helped the innovation in the machine travel.

A useful offer answers six questions. What outcome is promised? What must the customer give up in money, time, effort, attention and flexibility? What evidence supports the promise? What happens if the promise disappoints? How quickly does value arrive? What must the seller do repeatedly to make the economics work?

The fifth question is frequently underestimated. Time to value can dominate feature depth. Software that saves ten hours a week after a six-month implementation competes differently from a narrower tool producing two hours of savings tomorrow. A complex medical device, industrial machine or enterprise system may require training and process change before its technical capability becomes customer value. Implementation is therefore part of the offer whenever it determines the outcome.

Proof performs different jobs. A demonstration shows capability. A trial lets the buyer create personal evidence. A warranty reallocates downside. A reference from a similar customer transfers some trust. Independent testing can reduce information asymmetry. None proves the future, but each changes the cost of being wrong.

Bundling changes the unit being compared. A machine with maintenance is a different commercial object from the same machine sold alone. A software subscription combining updates and support converts uncertain future purchases into an ongoing relationship. Unbundling can reveal a cheaper core for customers who do not value extras. Both can increase clarity or concealment depending on how they are designed.

Offers also select customers. Generous free tiers may maximise sign-ups while attracting users with little reason to pay. Long contracts can improve revenue visibility while deterring buyers who fear lock-in. White-glove service can raise conversion and simultaneously destroy margin. Every reduction in customer sacrifice has a cost somewhere, so the offer must work for both sides over time.

The strongest offers reduce the buyer's most important uncertainty without creating obligations the seller cannot honour. They make a valuable product easier to choose and easier to realise.

One consequence is that conversion rate can be an incomplete measure of offer quality. A looser guarantee, deeper discount or easier sign-up may increase acceptance while worsening returns, fraud, support load or later churn. Offer design should therefore be judged by what happens after acceptance as well as by how many people accept.

4. Price Is Architecture, Not Arithmetic

Price is the most explicit term in the exchange, which is why it is often mistaken for a number. In practice it is a structure.

Cost-plus pricing begins with the seller's economics. Calculate cost, add a margin, obtain a price. That protects against selling below cost if the cost calculation is sound. It does not answer what the customer values or what the alternatives imply. Two services can cost the supplier the same amount while creating radically different savings for buyers. Cost places pressure on the seller's floor. It does not reveal the buyer's ceiling.

Value-based pricing starts from the customer's next-best alternative and the improvement the offer creates. What happens without us? What money, time, risk or opportunity cost does that alternative impose? How much of the gain can the seller capture while leaving the customer better off? The answer is rarely exact, but the question disciplines the design.

Price can also convey information when quality is hard to judge before purchase. Research has long found conditions in which buyers use price, brand or store cues when forming quality judgements. The effect is not universal. An implausibly high price on an unknown product can signal poor judgement instead of quality. Price works with category expectations, proof, reputation and channel.

One posted price can hide large differences in willingness to pay. Firms create versions, bundles, memberships, time restrictions, quantity breaks and other price fences so that customers with different needs select different bargains. Airline flexibility is an obvious example: two passengers can occupy neighbouring seats while paying different prices because the terms around change, timing and commitment differ.

This can expand access and improve efficiency. It can also become extraction. A price structure is easier to defend when the differences are intelligible, the buyer can make a meaningful choice and essential costs are not hidden until late in the process. It becomes harder to defend when complexity is used to exploit inattention, urgency or lack of alternatives.

Discounting deserves the same causal discipline as advertising. A promotion can accelerate trial, clear inventory or solve a timing problem. It can also pull purchases forward, alter reference prices and attract buyers whose loyalty is to the deal. An observed sales spike does not tell you which mechanism occurred. The useful question is what behaviour changed after the promotion ended.

For subscription businesses, the headline price is only one part of the architecture. Billing frequency, minimum term, usage limits, overage, upgrades, downgrades, renewal and cancellation can matter as much. For marketplaces, one side may be subsidised so that participation by the other side becomes more valuable. For industrial contracts, payment timing, service guarantees and risk sharing may dominate the sticker price.

Pricing therefore allocates value, segments customers and changes behaviour. It should be designed with the offer and channel, not handed over at the end as arithmetic.

Fairness matters because customers judge prices relationally. A price increase can be accepted when the reason is intelligible and rejected when it appears opportunistic, even if the economic value of the product has not changed. Firms therefore manage a social interpretation as well as a revenue equation. Hidden fees and surprise charges are especially corrosive because they change the bargain after the customer has invested attention or effort.

5. Channels Are Institutions With Power

A channel is often drawn as a line: producer, distributor, retailer, customer. The line hides the interesting part.

Channels perform work. They aggregate assortment, break bulk, hold inventory, provide local access, finance transactions, explain products, reduce search costs, deliver service, collect information and sometimes carry risk. A manufacturer can remove an intermediary from the diagram without removing these functions. Somebody still has to perform them and be paid.

This is why “going direct” is not automatically cheaper. A producer that leaves a wholesaler may gain margin and customer data, then discover that it has acquired warehousing, fulfilment, returns, support, demand generation and credit risk. Integration is attractive when control over those functions creates enough value to justify the cost and complexity. Independence is attractive when specialised intermediaries perform them better or create reach that would be expensive to reproduce.

Digital channels did not abolish intermediation. They created new intermediaries with extraordinary control over discovery. Search engines, app stores, social platforms and marketplaces can determine ranking, access, payment rules, fees, reviews and data. The channel can become the market's front door.

That produces power. A marketplace that delivers most of a seller's demand can change commission or ranking rules in ways the seller cannot easily resist. A retailer with scarce shelf space can demand promotional support. A franchise system can create conflict over local pricing or territory. Channel design is therefore partly a governance problem: who controls which decision, who owns the customer information, who bears inventory and service risk, and what happens when incentives diverge?

The channel also changes the proposition. Luxury goods sold through a discount environment carry different signals from the same goods sold through controlled distribution. Complex equipment sold through an expert integrator can include reassurance and implementation that a self-service website cannot. Convenience stores support a different basket and urgency from warehouse clubs. Distribution is part of customer experience because availability, advice, delay and returns are part of value.

Omnichannel strategies make the trade-offs more obvious. Customers may research in one place, inspect in another, purchase in a third and seek service in a fourth. Firms want continuity. Channel partners may want credit for the transaction. The customer does not care which internal budget owns the journey; the organisation often cares intensely.

A channel should therefore be evaluated on more than reach. Ask about contribution after fees and fulfilment, control of price and presentation, customer data, service quality, dependency, conflict, reversibility and strategic learning. A high-volume channel can be a poor business if it attracts low-value customers, compresses margin and prevents the seller from learning why people buy.

Channels do not merely carry demand. They shape what can be found, compared, trusted and bought. Choosing one is a choice about economics and power.

This is why distribution can become a competitive advantage in its own right. A product that is technically comparable to rivals can win because it is available where the need occurs, because replacement is immediate or because a trusted intermediary already knows how to install it. Availability is not glamorous, but unavailable value is commercially close to no value at all.

6. Growth Is Acquisition Plus What Happens Afterwards

The first purchase is psychologically satisfying because it is visible. Revenue appears, a conversion is recorded and somebody can take credit. Economically, it may be the middle of the story.

Acquisition quality varies. Two campaigns can produce the same number of customers with different repeat rates, support costs, returns, payment failures and willingness to recommend. Two channels can deliver identical first-order revenue while one produces customers who remain profitable and the other produces promotion hunters who vanish when the offer normalises. The source of a customer can change the future value of that customer.

That is why customer journeys matter. Awareness and consideration receive attention because marketers can intervene there, but onboarding, delivery, first use, service, renewal, cross-sell, cancellation and win-back all change the economics. A promise made before purchase becomes evidence afterwards. The experience either confirms the position or teaches the customer that the marketing was exaggerated.

Retention is valuable when a continuing relationship creates continuing value. It is not automatically valuable. Reinartz and Kumar's work challenged the easy assumption that long-life customers are always more profitable. Some demand service, wait for deals or generate little margin. Some short-lived customers can be highly profitable. Tenure is a behaviour, not a verdict.

The same caution applies to loyalty. In many repeat-purchase categories, larger brands tend to have both more buyers and somewhat higher loyalty, a pattern associated with the double-jeopardy literature. This is one reason retention cannot be treated as a separate magic lever detached from acquisition and market penetration. The exact pattern varies by context, but the broader lesson is useful: customer-base structure constrains what loyalty metrics mean.

A practical growth model follows cohorts. Group customers by acquisition period, channel, offer, segment or other meaningful condition and observe what happens later. Does a discount cohort reorder at full price? Do marketplace customers return directly? Do customers acquired through referrals cost less to support? Cohorts reconnect the first transaction with its consequences.

Lifetime value is useful when it is treated as a model rather than a fact. Future revenue, margin, retention and service cost must be estimated, often under changing conditions. A precise-looking number can hide fragile assumptions. The purpose is to force time into the decision, not to pretend the future is known.

Healthy retention also has an ethical test. A customer who stays because cancelling is obscure or costly is behaviourally retained but not necessarily well served. Regulators and researchers use terms such as dark patterns and online choice architecture for design practices that can steer or trap users in ways that benefit the firm at the consumer's expense. The commercial metric alone cannot tell the difference.

Marketing therefore owns more than the top of a funnel. It must care whether the promise survives contact with the product, whether the acquired customer becomes economically worthwhile and whether staying remains a voluntary choice.

The deeper lesson is that growth has composition. Ten thousand customers are not one thing. They arrived with different expectations, margins, service burdens and probabilities of returning. Aggregate growth can hide deterioration in that composition for a long time. Cohort analysis is useful because it makes the mixture visible before the average finally breaks.

7. Measurement Is a Model of the Market

Marketing measurement becomes dangerous when numbers are treated as observations without assumptions.

Suppose a customer sees an advert, searches for the brand and buys. An attribution system can assign the sale to the advert, the search click, the final website visit or some weighted combination. Each rule describes where credit goes. None by itself answers the causal question: would this customer have bought anyway?

Incrementality asks that counterfactual. The ideal comparison is between outcomes with the marketing activity and outcomes without it for otherwise comparable customers. Randomised experiments, geographic tests, holdouts and carefully designed quasi-experiments can get closer to that question than ordinary attribution. They are not effortless. Effects can be small relative to noisy purchasing behaviour, experiments can interfere with other activity and results can depend on time horizon and population.

Large-scale research on digital advertising has demonstrated the size of the problem. Gordon, Zettelmeyer, Bhargava and Chapsky compared observational methods with results from 15 Facebook advertising experiments involving hundreds of millions of user-experiment observations. The observational estimates often failed to recover the experimental effect reliably. Lewis and Rao showed a related difficulty: even large experiments can struggle to estimate advertising returns precisely because the incremental effect is small relative to the variation in customer spending.

The lesson is not that measurement is hopeless. It is that more tracking is not identical to more causal knowledge.

A useful measurement system begins with the decision it is supposed to inform. If the question is whether to increase spending on a channel, the relevant outcome may be incremental contribution over a suitable period, not attributed revenue this week. If the question is whether a positioning change improved fit, conversion may matter alongside win-loss reasons, retention and price sensitivity. If the question is whether a new segment deserves investment, the answer may require cohort economics and service cost.

Metrics also change behaviour inside firms. What receives a target receives attention. A team measured on leads can produce more low-quality leads. A retention team measured only on churn can make cancellation harder. A marketplace seller measured on ranking can sacrifice margin to keep velocity high. These are not failures of arithmetic. They are failures to model the system around the metric.

This closes the loop begun with market definition. Marketing starts by deciding which customers and decisions matter. Measurement later feeds evidence back into that choice. If the evidence is biased towards what is immediate, attributable or easy to count, the measurement system can silently retarget the business towards cheap converters, deal seekers or customers trapped by friction. The dashboard starts drawing the market boundary.

Good marketing measurement therefore needs three layers. Descriptive metrics tell what happened. Diagnostic analysis proposes why. Causal evidence tests whether the proposed intervention changed the outcome. None should be asked to do the work of the others.

The aim is not a perfect dashboard. It is a learning system capable of disagreeing with the organisation's favourite story.

That requires deciding in advance what result would change a decision. Otherwise every disappointing test can be explained away after the fact and every favourable correlation can be promoted to proof. Measurement becomes useful when it has consequences: stop the activity, change the target, redesign the offer, alter the channel or admit that the assumed mechanism was wrong.

How It Actually Works

Start with the buying situation

A workable marketing process begins with a concrete decision, not with a channel calendar.

Take a company selling scheduling software to clinics. “Healthcare” is not yet a useful market. The buyer may be a clinic owner replacing paper diaries, an operations manager replacing another software system or a hospital procurement team standardising across sites. The users may be receptionists and clinicians. Patients may experience the result without participating in the purchase. Each situation has different alternatives, risks, evidence requirements and routes to approval.

Write the decision in plain language: when this customer faces this situation, what are they trying to make happen, what do they do now, what blocks change and who else influences the choice? That sentence becomes the unit around which research, segmentation and positioning can be organised.

Build evidence from several angles

Begin with behaviour already occurring. Look at transaction data, lost deals, complaints, searches, switching, support requests, workarounds and the path customers take before buying. Then talk to people. Ask about the last real decision rather than a hypothetical future one. Reconstruct what triggered it, which alternatives appeared, who became involved, what delayed action and what finally made the choice feel safe enough.

Do not demand that customers design the solution. They are usually better witnesses to their constraints than forecasters of a product they have never used. Concept tests, surveys and purchase-intention measures can add evidence, but they should not be allowed to overrule contradictory behaviour merely because the numbers look cleaner.

Competitor research belongs here too. The relevant competitors are whatever customers use as alternatives. Record how each one creates value, where it is weak, what it costs in the broad sense and which customers it appears designed for. This is not a spreadsheet exercise performed for completeness. It tells you where a meaningful position might exist.

Choose the segment you can serve differently

Now partition the market only where the differences change action. One clinic may need rapid installation because its existing system is failing. Another may value integration and governance because switching risk is higher than delay. If those needs demand different offers, proof, service or channels, the distinction is commercially useful.

Assess attractiveness from both sides. Is the problem important enough for customers to act? Is the group reachable? Can the organisation deliver the required value better than the alternatives? Is the contribution sufficient after service and channel costs? Can the position survive likely competitive response? A large population with weak urgency can be less attractive than a small group with a severe, frequent problem.

Then state who is not the target. This prevents later pressure from turning every request into a feature and every lead into a customer.

Position against the real alternative

Write the comparison before writing the message. What do target customers use or do now? What do they believe about that alternative? Which capabilities of the offer create outcomes they value? Which evidence supports those outcomes? Which trade-offs follow?

A useful internal positioning statement can be ugly. Its job is clarity, not publication. It should identify the target, the relevant category or alternative, the valuable difference and the reason to believe it. From there, product priorities, proof and communication can be aligned.

If the claim requires caveats that customers would regard as disqualifying, the position is not ready. If the company must hide obvious weaknesses to make the sentence work, the position is aspirational rather than real.

Construct the offer around adoption

Now design the complete bargain. The product may need to change, but many adoption barriers sit around it: implementation effort, switching risk, contract length, uncertain support, payment timing, proof or time to value.

List the sacrifices explicitly. Money is one column. Add time, learning, internal approval, effort, uncertainty, loss of flexibility and reputational risk for the buyer. Then decide which sacrifice the seller can remove, absorb, insure or make reversible.

A trial works when personal experience is the missing evidence. A guarantee works when downside risk is the barrier and the seller can credibly carry it. Implementation support works when value is delayed by setup. Financing works when timing of cash, rather than total value, blocks the decision. Do not add every reassurance. Add the one that changes the choice.

Design price with the offer

Estimate the economics from three directions. The cost structure tells you what the seller must recover. The alternatives tell you what the customer is already sacrificing. Research and observed behaviour tell you how willingness to pay varies across segments and use cases.

Then choose the architecture: one price, tiers, bundles, usage, subscription, transaction fee, membership, service contract or another structure. The structure should correspond to how value is received where practical, while remaining understandable enough to compare.

Test the whole arrangement, not merely the number. A lower entry price with expensive required extras may increase clicks and reduce trust. A high price with clear implementation and risk transfer may be easier to approve than a cheaper but uncertain project.

Choose who performs the channel functions

Map the work required between production and use: discovery, assortment, explanation, inventory, financing, fulfilment, payment, setup, service, returns and information. Then decide which functions the firm will perform and which will be performed by distributors, retailers, platforms, agents, integrators or marketplaces.

For each route, calculate contribution after channel fees and the costs the firm still carries. Record who owns presentation, customer data, pricing discretion and service quality. Estimate dependency. A channel that produces rapid growth can still be strategically weak if losing access would remove most demand overnight.

Channel conflict is normal when routes overlap. A direct website may undercut retailers. A manufacturer may generate leads that distributors close. A marketplace may compete with its own sellers. Governance therefore needs rules about territory, pricing, lead ownership, service and data before conflict becomes the strategy.

Create demand and capture demand separately

Some marketing reaches people already looking for a solution. Search, marketplaces, comparison sites and sales enquiries often capture existing demand. Other activity builds memory, teaches a category or makes a problem salient before the buyer is ready. The distinction is imperfect, but useful.

If every activity is judged by immediate conversion, demand creation will usually look inefficient because many people influenced today buy later or elsewhere. If every brand-building activity is excused from measurement because its effects are long term, waste receives a comfortable hiding place. Match the evaluation period and method to the mechanism being claimed.

Advertising belongs here as one instrument among others. So do public relations, content, events, partnerships, referrals, merchandising and direct communication. The choice of instrument follows from where the customer can be reached and what needs to change in the decision. The deeper craft of paid communication belongs to advertising; marketing's job is to decide what role communication must perform inside the system.

Make the first purchase produce evidence

After purchase, track whether customers reach the promised value. Activation, implementation time, usage, repeat purchase, service contact, returns, renewal and cancellation can all reveal mismatch between the position and reality.

Analyse by cohort rather than averaging everything together. Customers acquired during a large promotion may behave differently from customers acquired through referrals. Enterprise customers onboarded by one partner may retain differently from those onboarded directly. Averages can conceal those mechanisms.

If retention is poor, avoid assuming the answer is a loyalty programme. The cause may be a wrong segment, oversold position, weak onboarding, product failure, service cost or an offer that attracted customers only at the introductory price. Fix the cause nearest to the mechanism.

Measure decisions, not activity

Finally, define which decisions require causal evidence. Descriptive reporting still matters: revenue, conversion, acquisition cost, repeat purchase, contribution and churn tell you what occurred. They should trigger questions rather than close them.

Where the decision is material, use tests that can estimate the counterfactual. Randomised holdouts are strong when feasible. Geographic or time-based designs can help when individual randomisation is impossible. Quasi-experimental methods can be useful when assumptions are credible. The method should follow the decision and the data, not fashion.

Then feed the result back to the beginning. Did the target segment behave as expected? Did the position attract the intended comparison? Did the offer remove the right barrier? Did the channel produce healthy cohorts? Did the activity cause incremental value? Change the weakest assumption rather than the easiest visible tactic.

Marketing rarely runs as a neat waterfall. Research continues while the offer changes; channels reveal new segments; pricing changes the customer mix; competitors react. The operating discipline is not rigid sequence. It is keeping the hypotheses connected enough that learning in one part changes the right part elsewhere.

One useful way to keep that connection is to maintain a small set of explicit commercial hypotheses. For example: independent clinics with more than five clinicians lose enough staff time to scheduling errors that they will pay for rapid implementation; they will compare the product mainly with their existing practice-management software; a migration service will remove the largest barrier; direct acquisition will be cheaper than reseller distribution at this scale; and customers who complete setup within a week will retain better. Each claim points to different evidence. If the first fails, changing an advert is irrelevant. If the fourth fails, the channel design is wrong.

The order of testing matters because some uncertainties dominate others. There is little value in optimising a subject line before establishing that the target problem is painful. There is little value in a sophisticated price experiment if implementation failure prevents customers reaching value. Early work should therefore reduce the largest strategic uncertainty first, even when the result is less tidy than a campaign metric.

Scale then creates a new problem. What worked when founders spoke directly to every customer may fail when acquisition broadens. A precise position can blur as new segments arrive. Service can become slower. The channel mix can change. Competitors can respond. Marketing at scale is partly the discipline of detecting when yesterday's fit is being diluted by today's growth.

That is also where demand creation becomes more important. In a small niche, the organisation may harvest buyers already searching for a solution. Growth eventually requires reaching people who are not currently in-market, teaching them a category or building enough memory that the brand is retrieved when the need appears later. This work is harder to connect to immediate revenue, which makes it vulnerable to short-term budgeting. Yet a business that only captures existing demand can exhaust the pool or become trapped in channels where competitors bid for the same ready-to-buy customers.

The reverse failure is equally common: celebrating awareness without commercial translation. Demand creation still needs a path into an offer, a price, a channel and an experience. A famous product that is difficult to buy, badly positioned at the moment of choice or unprofitable after fulfilment has not solved the marketing problem.

Competitive response should also be treated as part of the operating system. A successful offer teaches rivals. They can copy features, undercut price, enter the channel, raise bids for the same traffic or attack the category definition. The stronger the success, the more likely the environment changes. Strategy is therefore not a one-time fit but a moving one. The organisation needs some source of advantage that is costly to imitate, whether capability, data, distribution, trust, scale, switching cost, community, operational speed or a set of reinforcing activities.

Finally, write down the stopping rules. If a segment does not reach a minimum contribution after a defined learning period, stop treating more acquisition as the answer. If a channel produces volume but unacceptable dependency, cap it. If a price test increases conversion but lowers contribution and retention, reject the apparent win. If an experiment cannot distinguish the effect that matters, admit uncertainty rather than converting noise into certainty.

The operating sequence is therefore iterative but not vague. Begin with a buying situation, build evidence, choose, position, construct the offer, price it, select the route, create and capture demand, observe the experience and test causes. Then return to the first assumptions with better information.

A final complication is organisational ownership. The decisions above rarely sit in one department. Product controls features and reliability. Finance influences price and acceptable payback. Operations determines delivery promises. Sales hears objections and can distort the customer mix if incentives reward any closed deal. Service sees the gap between promise and experience. Marketing becomes ineffective when it collects these signals but lacks any mechanism for changing the system they describe.

The remedy is not to put every function under a chief marketing officer. It is to make the market-facing hypotheses shared. If the position depends on two-day delivery, operations needs to treat that as part of the value proposition. If a premium price depends on lower implementation risk, service quality is not an afterthought. If sales repeatedly wins customers outside the target and those customers churn, the sales incentive and the marketing strategy are in conflict. The boundary of the marketing department matters less than whether the organisation can coordinate around the chosen exchange.

This is also why annual marketing plans age badly in volatile markets. A plan can lock in spend and channel commitments after the assumptions that justified them have changed. Better systems separate relatively durable choices, such as target customer and core position, from adjustable choices, such as campaign mix, promotional timing and channel budget. The durable choices should change only with evidence strong enough to justify disruption. The adjustable choices should change faster.

There is a useful hierarchy of failure. If nobody cares about the problem, the market hypothesis is weak. If the target cares but does not understand why this offer fits, positioning is weak. If the position is clear but people still refuse, the offer, price, proof or risk may be wrong. If people buy but do not reach value, delivery is weak. If customers succeed but growth stalls, reach or availability may be weak. If the business cannot tell which of these is happening, measurement is weak.

That hierarchy prevents a common reflex: asking the most visible team to work harder. More traffic does not cure poor retention. More content does not cure a bad price architecture. More sales pressure does not cure an offer that produces buyer's remorse. Marketing works when the diagnosis points to the mechanism rather than to the department currently under pressure.

The process ends where it began, with choice. A company cannot optimise every customer, every channel and every time horizon at once. It must decide whether it values reach or control, rapid adoption or high qualification, standardisation or service, current contribution or a longer payback. Those trade-offs are not defects to be eliminated by better analytics. They are the substance of strategy. Marketing makes them visible in the language of customers and markets.

Good teams therefore keep a small number of measures that correspond to those choices and resist the temptation to manage by a hundred convenient indicators. The purpose of a metric is to improve a decision. Once a measure no longer changes a decision, or once people learn to game it, its apparent precision becomes a liability. The system should be able to retire metrics as well as add them.

How we know

Marketing evidence is unusually mixed because the subject sits between behaviour, organisations and competitive markets. Transactions and experiments can establish some effects strongly. Surveys, interviews and observation explain motives and constraints imperfectly but usefully. Historical cases show mechanisms without proving that the same result will recur elsewhere.

Several claims in this book are therefore conditional rather than laws. Price can signal quality under some conditions. More choice can help or hinder depending on the decision. Long tenure does not guarantee profitability. Channel integration can improve control and still raise cost. Digital tracking can describe detailed paths while remaining weak evidence of causation.

The strongest generalisations are mechanisms rather than recipes: buyers compare alternatives; terms change risk; intermediaries perform functions and possess incentives; retention has economics; attribution differs from incrementality. Where a result depends heavily on category, customer or time horizon, the book says so rather than turning an average effect into a universal rule.

What People Get Wrong

“Marketing is promotion”

Promotion is the part of marketing that interrupts you, so it becomes the whole discipline in public memory. Companies reinforce the mistake by sending every request for growth, launches or social posts to a department called marketing while product, price and distribution sit elsewhere.

The four Ps were meant to prevent this collapse. Product, price, place and promotion described coordinated managerial choices, and even that framework omitted much of the customer journey now visible in services and subscriptions. An advert can introduce an offer. It cannot decide which market is worth serving, remove adoption risk, build a workable price or place stock where the customer can buy it.

The error also distorts budgets. Communication is funded as a campaign while research, onboarding, pricing work and channel support are treated as overhead. The organisation then demands a measurable return from the visible slice and ignores the decisions that made the return possible or impossible.

The distinction matters because promotion magnifies the system it receives. If the position is vague, the offer weak or delivery poor, more attention accelerates waste and disappointment. Ask what must change in the exchange before asking how to announce it.

“A better product sells itself”

The claim flatters builders because it makes commercial success a verdict on objective quality. If the product is good enough, the market will notice. Failure can then be blamed on irrational customers or inadequate exposure.

Products do not enter a laboratory comparison. Buyers face categories, habits, budgets, switching costs, risk and incomplete information. They may never encounter the offer, may place it in the wrong frame or may value a different attribute from the one the producer improved. A technically superior product can lose to an adequate alternative that is familiar, available and easier to trust.

Product teams can make the error in reverse by measuring quality through internal difficulty. A feature that required heroic engineering is not automatically valuable, while a mundane reduction in setup time may decide the sale. Customer value has no duty to mirror production effort. Quality still matters, and clever marketing cannot rescue a product that fails in use. Delivered value is the foundation of retention and evidence. Quality must be legible inside a decision. It needs a target, comparison, offer, price, channel and proof. Superiority that the customer cannot perceive or receive remains private achievement.

“Customers know what they want”

Customers know a great deal about their problems, routines, frustrations and previous choices. They are less reliable as designers of unfamiliar solutions or prophets of their own future behaviour. The slogan persists because asking feels scientific and produces quotable answers.

Research on purchase intentions finds a real relationship with later behaviour, but an imperfect and conditional one. Hypothetical approval costs nothing. Actual purchase competes with price, effort, timing, approval and alternatives. People also describe needs through the products they already know, which can narrow the solution before the problem is understood.

Novelty changes the question. People could not have requested a product category they had never encountered, but they could describe the delays, costs and awkward workarounds the category later removed. Research should be humble about solutions and precise about problems.

The opposite slogan, that customers have no idea what they want, is equally lazy. It gives the firm permission to ignore evidence. Ask customers about specific past behaviour, consequences and trade-offs. Observe what they use and pay for. Test proposed solutions through commitments. Treat words as evidence about meaning and action as evidence about choice.

“More choice is always better”

More options can serve diverse needs, increase the chance of a good match and reduce search across sellers. The mistake is assuming that these benefits rise without limit and in every context. Choice overload became famous through striking experiments, then hardened into a universal law. A 2010 meta-analysis of 63 experimental conditions found an average effect close to zero and substantial variation. Larger assortments can help experienced buyers with clear preferences. They can hinder buyers when options are unfamiliar, hard to compare and similar enough to make regret likely.

Assortment also serves the seller. Extra versions can occupy shelf space, defend price points or help a channel serve different buyers. Those aims may be rational, yet they should not be confused with customer benefit when the distinctions are hard to use.

The useful response is not to cut every range. It is to design choice. Use clear categories, meaningful versions, defaults that can be defended and comparison attributes tied to the decision. Variety creates value when it maps to real differences in need. It creates work when the organisation has multiplied options without helping customers understand why they exist.

“Lower prices create loyalty”

A discount produces action, and action looks like attachment. Sales rise, new customers appear and the promotion receives credit. The story is persuasive because the result is immediate while the customer's motive remains invisible.

Price promotions can prompt trial, move stock and change purchase timing. Evidence from consumer categories has often found little permanent sales effect once the promotion ends, and repeated discounts can increase attention to price. The customer may have learned when to buy cheaply rather than why to prefer the offer.

Discounts can also damage reference points. Once customers have repeatedly seen a product at twenty per cent off, the standard price may feel inflated even when nothing about cost or value changed. A temporary tactic has rewritten the expected bargain.

Loyalty built on value can survive a normal price. Loyalty built on a deal often waits for the next deal. That does not make discounting wrong. It makes the objective important. Use a promotion to solve a defined problem, then measure repeat contribution after the offer ends. If demand vanishes at the list price, the promotion exposed the position instead of strengthening it.

“Loyal customers are always the most profitable”

Retention is cheaper than acquisition is repeated as though it were a physical law. It can be true when repeat customers require less persuasion, buy more and cost less to serve. It fails when tenure is used as a substitute for economics.

Research on long-life customers in noncontractual settings found that longevity did not guarantee profitability. Some established customers bought little, demanded heavy service or expected lower prices. A customer can also remain because switching is difficult, which creates duration without advocacy or healthy value. Contractual businesses add another trap. Low churn can reflect annual commitments, cancellation friction or customer neglect rather than satisfaction. Renewal becomes informative only when customers had a fair opportunity to leave and continued use shows that value survived the signature.

Measure contribution by cohort and behaviour, not affection alone. Retain customers for whom the exchange works, repair experiences that destroy good fit and allow poor-fit relationships to end cleanly. Growth also requires penetration. Smaller brands commonly have fewer buyers and slightly lower repeat rates, so a loyalty-only strategy can perfect a base that is too small.

“Digital marketing tells you exactly what worked”

Digital systems record a detailed sequence, which creates the impression of causality. A customer saw a post, clicked a search result and bought. The dashboard awards credit, often to the last visible touchpoint. The missing fact is what the customer would have done without each touch. Attribution follows the observed path. Incrementality requires a counterfactual. Large field experiments have shown that observational estimates can differ materially from experimental estimates, and even experiments can leave wide uncertainty when purchase behaviour is noisy and effects are small.

Platform reports also face an incentive problem. The system selling an impression may be the system reporting its value, and each platform can claim the same sale under its own attribution rule. Adding the claims together does not create more customers.

Digital data remains useful for diagnosis, operations and hypothesis generation. Use holdouts, randomisation or credible comparisons when the decision requires causal proof. Match the measurement horizon to the effect. A click is evidence that a click occurred. It is not automatic proof that demand, profit or a customer relationship was created.

Use It

Find the buying situation, not the demographic

When somebody presents a target audience, ask for the decision underneath it. What happened before the customer began looking? What is being compared? Who pays, uses, approves and objects? What cost does delay create?

Demographics may help locate people, but they rarely tell you which trade-off will move them. Replace broad descriptions such as small business owners or affluent professionals with a situation precise enough to change the offer. A firm hiring its first employee faces a different payroll decision from a fifty-person company replacing a failed system, even though both are small businesses.

Then test whether the segment earns a different position, price, channel or service. If it changes none of those, the label is unlikely to improve the decision. Use the market boundary as an operating rule, not a slide. When a request arrives from outside it, decide consciously whether to refuse, build a separate version or change the target.

State the alternative you are replacing

Before writing a value proposition, complete one sentence: the customer would otherwise do this.

The answer may be a competitor, a manual process, an internal hire, a familiar workaround or nothing. Name it in practical terms. If the alternative is doing nothing, explain what event makes inaction newly costly. If the alternative is an established supplier, identify what the customer would lose by switching as well as what they might gain.

Now compare the offer on the attributes that decide the choice. Avoid lists of benefits with no reference point. Faster than what, easier for whom, safer by which evidence? A useful position makes one difference clear without pretending every other attribute is superior.

This lens also improves competitive research. Do not ask only what rivals say. Ask why customers continue to choose them. Familiarity, availability, integration and low perceived risk can defeat a better feature because they belong to the decision the buyer is making.

Price the whole customer sacrifice

Money is one part of price. Add the time to evaluate, the effort to switch, the risk of failure, the attention required, the internal approval and the commitments hidden in the terms.

Then inspect how the offer redistributes those sacrifices. A trial reduces commitment but may increase setup effort. A managed service charges more money and removes labour. Annual billing can lower the rate while increasing lock-in. A marketplace may make comparison safer while narrowing direct support.

This approach prevents a common mistake: lowering the monetary price when the real barrier sits elsewhere. If buyers fear migration, a discount leaves the fear intact. If they lack proof, a guarantee or reference may change the decision more efficiently. If implementation takes months, staged delivery may create value sooner.

On the seller side, include channel fees, support and failure, not merely production cost. A price is sustainable only when it funds the experience promised to the target customer.

Trace the channel power as well as the channel tax

For every route to market, write down what the channel contributes and what it takes.

Contribution may include traffic, trust, assortment, local stock, payment, demonstration, finance, delivery, support and returns. The cost may include margin, data, control, dependence, price restrictions and distance from the customer. A high commission can be economical when it replaces expensive work. A low commission can be costly when the channel owns demand and makes the seller interchangeable.

Follow the channel backwards into the offer. Does packaging suit transport? Can the margin support service? Does the channel understand the position? Are returns and guarantees workable? Then follow it forwards into power. Who can change the terms, hide the offer, favour a rival or prevent the seller from taking the relationship elsewhere?

Do this before celebrating reach. A channel can increase revenue while worsening contribution and weakening the firm's ability to learn. Growth through dependence is still dependence.

Separate demand creation from demand capture

Classify activity by the state of the customer when it reaches them.

Demand capture meets a buyer already close to action: a search result, a stocked shelf, a comparison page, a referral request or a sales conversation. Demand creation makes the problem, category or offer easier to notice and remember before that moment. Both matter. They produce evidence on different schedules.

Capture often wins short-term reporting because the sale follows quickly and the trail is visible. That can cause firms to overfund the bottom of the journey while relying on past reputation, existing need or competitors to create the demand being harvested. Creation can be wasted too, especially when the offer is unavailable or the next step is unclear.

Measure each against its job. Ask whether capture was incremental and profitable. Ask whether creation changed awareness, consideration, search, price response or later sales across a suitable horizon. Do not force every effect into the same attribution window.

Build a causal learning loop that can disagree with you

Write the important assumptions before the result arrives. The target has this problem. This alternative is unsatisfactory. This proof will reduce risk. This price will produce enough contribution. This channel will reach the customer without destroying the experience.

Attach evidence and a decision threshold to each. Interviews may test language and process. Transactions test willingness to act. Cohorts test later value. Experiments test causal changes where a credible comparison is possible. Service and sales evidence test where the promise breaks.

Then protect disconfirming evidence. Teams naturally explain away the customers who did not convert, the channel that produced low-quality demand and the cohort that churned after an impressive launch. A learning system should make those contradictions visible before a success story hardens around the aggregate.

Do not change everything at once. Preserve enough consistency to let the market learn the position and to distinguish signal from noise. The loop must be capable of disagreement without becoming addicted to novelty.

The limits

Marketing can improve the fit between an offer and a choice. It cannot create unlimited need, abolish competition or guarantee that people behave as research predicts. Markets change while firms study them. Buyers have conflicting motives. Competitors respond. Evidence gathered from one period, channel or segment may fail elsewhere.

The discipline also has an ethical limit that commercial metrics do not enforce. A design can raise conversion by hiding a fee, manufacturing urgency, making cancellation exhausting or exploiting private information. Official reports in the United States and United Kingdom have documented these dark patterns because a completed transaction does not prove informed or voluntary choice. The customer can click while being misled.

Marketing cannot decide by itself which exchanges should exist. A profitable offer can impose costs on workers, communities, public institutions or the environment that neither buyer nor seller pays. It can deepen harmful consumption or concentrate power in a channel. Value for the parties is not the whole social account.

Finally, marketing language can overstate control. Success is often credited to strategy after the fact, while timing, luck, capital and competitor mistakes vanish from the story. Use models as instruments for better decisions, not as proof that the market has become obedient.

The one thing to keep

Keep the exchange in view.

When somebody says marketing, do not picture the advert first. Picture a person facing alternatives. Ask what they are trying to change, what they must give up, why this offer belongs in the comparison, how it reaches them, what happens after purchase and what evidence would show that both sides received value.

That view joins decisions organisations prefer to separate. A feature alters the position. A price alters the customer. A channel alters the product and the margin. A promotion alters expectations. A service failure alters the cost of the next sale. A metric alters what the organisation notices and therefore what it becomes.

In 1962 Rolls-Royce did not make the Viper engine valuable through a louder claim. It changed the bargain around the engine so that operator and manufacturer were paid by the same outcome: hours of reliable flight. The offer, price, service and relationship became one system.

That is the permanent shift. Marketing is not the work of making people want what has already been decided. It is the work of deciding what exchange is worth building, for whom, against what alternative and through which connected choices. Communication comes after there is something coherent to communicate.

Once you see marketing as the design of the exchange, bad strategy becomes harder to hide behind busy activity, and good strategy becomes visible in places that never looked like marketing at all.

Terms

Market. A set of buyers and situations organised around a decision and its alternatives. A market is defined for a purpose; change the decision and its useful boundary may change.

Category. The mental and commercial group in which an offer is compared. Category membership supplies expectations, competitors and buying rules, which is why choosing the frame affects the meaning of features.

Segment. A group of customers who differ in a way that justifies a different marketing choice. A segment matters when it changes position, offer, price, channel, service or communication.

Target market. The segment whose decision governs the design of the marketing system. Other customers may buy, but the primary target receives priority when preferences and trade-offs conflict.

Need state. The condition or situation that creates a desired change, such as urgency, uncertainty or overload. The same person can enter different need states and make different choices.

Job to be done. A way of describing the progress a customer seeks in a particular circumstance. It directs attention towards the problem and context rather than the product category alone.

Buyer. The person or organisation making or approving the purchase. The buyer may differ from the user, payer and beneficiary, which creates separate requirements and sources of resistance.

User. The person who experiences the product or service in practice. User value drives adoption and retention even when another party controls the budget and signs the agreement.

Value proposition. A concise account of the value offered to a target customer relative to alternatives. It should identify a meaningful outcome and the reason the organisation can deliver it credibly.

Positioning. The deliberate place an offer occupies in a buyer's comparison. Positioning links target, frame, parity, difference and proof, then guides product, price, channel and communication choices.

Frame of reference. The category or alternative set used to interpret an offer. The frame makes unfamiliar features intelligible but also imposes minimum expectations the offer must satisfy.

Point of parity. A capability required to be considered a credible member of a category or a viable alternative. Parity earns entry to the comparison before difference can decide it.

Point of difference. A valuable distinction the target customer can perceive and the organisation can support. Strong differences arise from capabilities and linked choices, not from an unsupported adjective.

Offer. The complete bargain placed before the customer: product, price, terms, proof, service, delivery, risk and timing. Changing the offer can change adoption without changing the underlying product.

Willingness to pay. The maximum sacrifice a buyer would accept for an offer in a given situation. It varies with alternatives, urgency, proof, budget, risk and the way the offer is framed.

Price fence. A rule that lets different customers access different prices or versions, such as timing, quantity, eligibility or flexibility. A defensible fence supports segmentation without arbitrary treatment.

Versioning. Designing different forms of an offer for customers with different needs or willingness to pay. Useful versions help buyers select themselves and fund service levels that differ meaningfully.

Bundle. Two or more products or services sold as one offer. Bundling can reduce search and transaction effort, raise total value or conceal components the customer would reject separately.

Channel. The path connecting an organisation to discovery, evaluation, purchase, delivery and support. A channel affects margin, control, data, availability, trust and the customer experience.

Intermediary. An independent party that performs channel functions between producer and customer. Retailers, distributors, marketplaces and agents earn a share by reducing work, risk or search for others.

Distribution. The system that makes an offer available in the required place, time, quantity and form. Distribution includes physical movement, inventory, assortment, access and ownership of channel relationships.

Reach. The number or proportion of relevant people exposed to an offer or communication. Reach describes potential contact, not attention, understanding, action or incremental commercial effect.

Penetration. The proportion of category buyers who purchase a brand or offer during a period. Penetration is distinct from purchase frequency and is often a major source of growth.

Acquisition. The process and cost of gaining a new customer. A low acquisition cost is useful only when the customer produces enough later contribution and the channel does not destroy the relationship.

Conversion. Movement from one defined stage to another, such as visit to purchase. Conversion rates diagnose a particular step; they do not by themselves prove demand creation, profit or fit.

Retention. Continued customer activity or relationship across a stated period. Retention can reflect value, habit, contract or friction, so the mechanism and economics matter more than duration alone.

Churn. The loss or inactivity of customers during a period. Churn should be separated by cause, cohort, target and channel because different departures require different responses.

Customer lifetime value. An estimate of the present value of future customer contribution after relevant costs. It is a model built from retention, margin, timing and behaviour, not a permanent customer trait.

Attribution. A rule for assigning observed credit among touchpoints before an outcome. Attribution describes a recorded path and depends heavily on which interactions the system can see.

Incrementality. The additional outcome caused by an activity compared with what would have happened without it. Incrementality requires a credible counterfactual and is the stronger basis for resource decisions.

Go Deeper

The overview: Philip Kotler, Kevin Lane Keller and Alexander Chernev, Marketing Management, 16th edition (Pearson, 2022). This is the large reference work behind much of modern marketing education. It covers market analysis, segmentation, positioning, offerings, pricing, channels, communication and customer value in one organised system. Use it to extend any part of this book and to see the standard vocabulary in context. The warning is size: it is a textbook built for a course, not a continuous narrative, so read by question rather than from page one. The chapters on market orientation and customer value are the best entry points before moving into the specialist sections.

Positioning: April Dunford, Obviously Awesome: How to Nail Product Positioning so Customers Get It, Buy It, Love It (2019). Dunford turns positioning from an abstract statement into a practical sequence beginning with competitive alternatives, distinctive capabilities, customer value and the segment that cares most. It is especially useful for technology and business-to-business offers whose value is real but hard to explain. The method is narrower than the whole discipline, which is its strength. Read it when an offer works but buyers place it in the wrong comparison. It is also a useful antidote to positioning exercises that begin with desired adjectives rather than evidence about alternatives.

Pricing: Thomas T. Nagle, John E. Hogan and Joseph Zale, The Strategy and Tactics of Pricing, 5th edition, New International Edition (Routledge, 2016). This is the place to go when cost-plus pricing and casual competitor matching stop being adequate. It explains value, willingness to pay, segmentation, price structures, communication and organisational implementation. The book is analytical without requiring advanced mathematics. Its examples and terminology are built for managers, though the treatment is dense enough that a real pricing problem makes the best reading companion. Work through it alongside one live offer, because the distinctions between price level, structure and communication become clearer when real numbers and customer alternatives are present.

Channels: Anne T. Coughlan, Erin Anderson, Louis W. Stern and Adel I. El-Ansary, Marketing Channels, 7th edition (Prentice Hall, 2006). This is the serious treatment of the part of marketing most popular books neglect. It explains why intermediaries exist, which service outputs customers require, how channel structures are chosen and why power and conflict emerge. Some examples and digital references have aged, but the underlying mechanisms remain useful. Read it when a direct, retail, distributor or marketplace decision looks like a margin comparison and is probably much more. Its account of channel functions helps identify the work that returns to the producer when an intermediary is removed.

Notes and Sources

The Whole Thing in One Page

Power-by-the-Hour. Rolls-Royce's official 2012 history dates the service concept to 1962 and describes a fixed-cost-per-flying-hour arrangement for the Viper engine on the de Havilland/Hawker Siddeley 125, including complete engine and accessory replacement. The company's account explicitly states that payment aligned the interests of manufacturer and operator around engines that performed well. The example is used to show a change in offer, price, risk and relationship, not to claim that outcome-based service began with Rolls-Royce.

The breadth of marketing. The American Marketing Association's current definition covers creating, communicating, delivering and exchanging offerings of value. Borden's 1964 marketing-mix article and McCarthy's 1960 textbook establish that promotion was one element in a broader managerial system. The manuscript uses a wider exchange model because services, subscriptions, customer journeys and data make the interdependence of the elements more visible.

Why You Should Care

Xerox 914. The Smithsonian National Museum of American History dates the machine's introduction to 1959 and describes its importance as the first successful commercial plain-paper copier. David Owen's history supplies the commercial context. Historical pricing varied across periods and accounts, so the narrative retains the stable point: leasing and usage-linked charging reduced the barrier of outright purchase and connected revenue to copying activity. No exact fee is required for the argument.

Choice architecture. The idea that comparison pages, defaults, renewals and rankings shape decisions is supported by the UK Competition and Markets Authority's 2022 evidence review of online choice architecture and the U.S. Federal Trade Commission's 2022 report on dark patterns. The text uses these sources for mechanisms and examples, not for legal advice.

The Core Ideas

Markets, segments and targets. Smith's 1956 article is the foundational source for treating segmentation and product differentiation as responses to heterogeneous demand. Kohli and Jaworski's 1990 account defines market orientation through organisation-wide intelligence generation, dissemination and response. Day's 1994 article develops market-sensing and customer-linking capabilities. These works support the claim that market choice is a continuing organisational process rather than a demographic exercise owned by one department.

Stated and revealed evidence. Morwitz's 1997 review supports the cautious treatment of purchase intentions: stated intentions are related to later behaviour but are imperfect and affected by context. The book does not claim that interviews are unreliable. It separates what interviews reveal well, such as language and experienced problems, from the stronger evidence required for future purchase and willingness to pay.

Positioning. Kotler, Keller and Chernev supply the standard structure of target, competitive frame, points of parity and points of difference. Dunford provides a practical alternative-first sequence. Porter's 1996 article supports the distinction between a slogan and a position sustained by trade-offs and fit among activities; Southwest Airlines and IKEA are his examples. The book uses them to illustrate linked systems, without claiming either company's current operation remains unchanged.

Offer design. Rolls-Royce and Xerox provide the principal concrete anchors. The offer model also draws on the customer-value and pricing literature: Zeithaml's 1988 synthesis treats perceived value as a relationship between benefits and sacrifices, while Nagle, Hogan and Zale analyse value, price structure and segmentation. Guarantees, trials and staged commitments are presented as mechanisms that redistribute risk, not as universally effective tactics.

Price. Zeithaml supports the treatment of price as both sacrifice and information. Rao and Monroe's 1989 review supports a conditional price-quality inference rather than the broad claim that high prices always signal quality. Hinterhuber distinguishes cost-based, competition-based and customer-value-based pricing and documents implementation resistance. Nagle, Hogan and Zale support willingness to pay, versioning, price fences and the need to connect price with value communication and costs.

Channels. Bucklin's theory explains why channel structures arise from discrepancies in quantity, assortment, timing and place. Coughlan and colleagues provide the main treatment of channel functions, service outputs, power and conflict. Anderson and Coughlan's semiconductor study supports the claim that integrated versus independent distribution is a governance and capability choice. Frazier supplies the treatment of dependence and channel coordination.

Customer journey and retention. Lemon and Verhoef organise customer experience across pre-purchase, purchase and post-purchase touchpoints and stress the role of multiple channels and partners. Gupta and Lehmann support the treatment of customers as assets whose future contribution must be estimated. Reinartz and Kumar's noncontractual customer data support the narrower correction that long tenure does not guarantee profitability. Ehrenberg, Goodhardt and Barwise support the double-jeopardy pattern: smaller brands commonly have fewer buyers and slightly lower repeat purchase.

Measurement. Lewis and Rao analyse 25 large digital advertising experiments and show why precise return estimates can remain difficult even at scale. Gordon and colleagues compare observational approaches with large Facebook field experiments and find material divergence. These sources support the distinction between attribution and incrementality. Binet and Field are used only for the risk of allowing immediate response measures to crowd out effects that develop over longer periods.

How It Actually Works

The marketing mix. Borden's original list was broader than the later four Ps. McCarthy's product, price, place and promotion structure remains useful as a memory aid. The book's criticism is directed at treating the elements as independent controls, not at the framework itself.

Research and segmentation. The operating sequence combines Smith, Kohli and Jaworski, Day, Morwitz and standard treatments in Kotler, Keller and Chernev. The bottom-up market-size test and the distinction between prevalence and consequence are managerial deductions from those mechanisms rather than claims tied to a single study.

Position, offer, price and route to market. Porter supports trade-offs and activity fit. Dunford supports mapping alternatives before writing a position. Zeithaml and Nagle and colleagues support customer sacrifice and pricing architecture. Bucklin and Coughlan and colleagues support the functional channel map. The manuscript's step sequence is an editorial synthesis designed for a one-hour mental model, not a claim that all organisations should follow a rigid linear process.

Cohorts and causal learning. Lemon and Verhoef support the full journey. Gupta and Lehmann support lifetime contribution. Lewis and Rao and Gordon and colleagues support holdouts and causal caution. Cohort analysis is presented as a way to preserve the link between acquisition conditions and later behaviour; no universal threshold is prescribed.

How we know. The limits paragraph reflects the mixed evidence base of marketing: field experiments, observational data, surveys, transactions, case studies and historical accounts answer different questions. The wording distinguishes strong recurring mechanisms from effects conditional on category, customer, competition and time.

What People Get Wrong

Marketing and product superiority. The first two corrections derive from the marketing-mix, market-orientation and positioning literature. The manuscript does not argue that communication or product quality is unimportant. It argues that neither can substitute for the complete exchange.

Customer knowledge. Morwitz supports the gap between stated intention and later behaviour. The correction deliberately rejects both extremes: asking customers to design the solution and dismissing customer evidence entirely.

Choice overload. Scheibehenne, Greifeneder and Todd's 2010 meta-analysis covers 63 conditions from 50 published and unpublished experiments. The mean effect was close to zero with substantial variation. This is why the text corrects the universal claim without denying that overload occurs under some conditions.

Discounting. Pauwels, Hanssens and Siddarth studied scanner-panel data for a perishable and a storable product. They found that permanent promotion effects were largely absent across category incidence, brand choice and purchase quantity in those settings. The manuscript narrows the conclusion to caution about durable loyalty and reference prices rather than claiming all promotions fail.

Loyalty and profit. Reinartz and Kumar's evidence shows that both short-life and long-life customers can be profitable and that simple tenure-profit assumptions fail. The book therefore treats retention as valuable when the contribution economics and experience support it.

Digital measurement. Gordon and colleagues provide the comparison between observational and experimental approaches. Lewis and Rao provide the precision problem. The dashboard and platform-incentive observations are logical consequences of attribution systems and do not allege misconduct by a named platform.

Use It

Ethical limits. The FTC report documents disguised advertising, hidden terms, difficult cancellation and practices that induce data disclosure. The CMA review covers online choice architecture that can pressure, deceive or impair choice. Both were checked on 9 August 2026. The book makes the broader ethical claim that conversion is not sufficient evidence of informed welfare. It does not summarise current law or compliance requirements.

External costs and control. The statement that buyer and seller value does not exhaust the social account is a general limit on the exchange model. It is included to prevent commercial success from being treated as proof of net social benefit.

Terms and Go Deeper

Definitions follow the usage established in the sources above. The four recommendations were verified against publisher, author or institutional records. Editions are stated in the Bibliography. Marketing Management is a broad reference; Obviously Awesome is a practical positioning guide; The Strategy and Tactics of Pricing treats price; and Marketing Channels treats distribution structure and governance.

Bibliography

Official and historical sources

American Marketing Association. “Definitions of Marketing.” Accessed 9 August 2026.

Competition and Markets Authority. Online Choice Architecture: How Digital Design Can Harm Competition and Consumers. London: CMA, 2022.

Federal Trade Commission. Bringing Dark Patterns to Light. Washington, DC: FTC, 2022.

National Museum of American History. “Xerox 914 Plain Paper Copier.” Smithsonian Institution. Accessed 9 August 2026.

Rolls-Royce. “Rolls-Royce Celebrates 50th Anniversary of Power-by-the-Hour.” Press release, 30 October 2012.

Original research and foundational articles

Anderson, Erin, and Anne T. Coughlan. “International Market Entry and Expansion via Independent or Integrated Channels of Distribution.” Journal of Marketing 51, no. 1 (1987): 71-82. DOI 10.1177/002224298705100106.

Borden, Neil H. “The Concept of the Marketing Mix.” Journal of Advertising Research 4, no. 2 (1964): 2-7. DOI 10.1080/00218499.1964.12519724.

Day, George S. “The Capabilities of Market-Driven Organizations.” Journal of Marketing 58, no. 4 (1994): 37-52. DOI 10.1177/002224299405800404.

Ehrenberg, Andrew S. C., Gerald J. Goodhardt, and T. Patrick Barwise. “Double Jeopardy Revisited.” Journal of Marketing 54, no. 3 (1990): 82-91. DOI 10.1177/002224299005400307.

Frazier, Gary L. “Organizing and Managing Channels of Distribution.” Journal of the Academy of Marketing Science 27, no. 2 (1999): 226-240. DOI 10.1177/0092070399272007.

Gordon, Brett R., Florian Zettelmeyer, Neha Bhargava, and Dan Chapsky. “A Comparison of Approaches to Advertising Measurement: Evidence from Big Field Experiments at Facebook.” Marketing Science 38, no. 2 (2019): 193-225. DOI 10.1287/mksc.2018.1135.

Gupta, Sunil, and Donald R. Lehmann. “Customers as Assets.” Journal of Interactive Marketing 17, no. 1 (2003): 9-24. DOI 10.1002/dir.10045.

Hinterhuber, Andreas. “Customer Value-Based Pricing Strategies: Why Companies Resist.” Journal of Business Strategy 29, no. 4 (2008): 41-50.

Kohli, Ajay K., and Bernard J. Jaworski. “Market Orientation: The Construct, Research Propositions, and Managerial Implications.” Journal of Marketing 54, no. 2 (1990): 1-18. DOI 10.1177/002224299005400201.

Lemon, Katherine N., and Peter C. Verhoef. “Understanding Customer Experience Throughout the Customer Journey.” Journal of Marketing 80, no. 6 (2016): 69-96. DOI 10.1509/jm.15.0420.

Lewis, Randall A., and Justin M. Rao. “The Unfavorable Economics of Measuring the Returns to Advertising.” Quarterly Journal of Economics 130, no. 4 (2015): 1941-1973. DOI 10.1093/qje/qjv023.

Morwitz, Vicki G. “Why Consumers Don't Always Accurately Predict Their Own Future Behavior.” Marketing Letters 8, no. 1 (1997): 57-70. DOI 10.1023/A:1007937327719.

Pauwels, Koen, Dominique M. Hanssens, and S. Siddarth. “The Long-Term Effects of Price Promotions on Category Incidence, Brand Choice, and Purchase Quantity.” Journal of Marketing Research 39, no. 4 (2002): 421-439. DOI 10.1509/jmkr.39.4.421.19114.

Porter, Michael E. “What Is Strategy?” Harvard Business Review 74, no. 6 (1996): 61-78.

Rao, Akshay R., and Kent B. Monroe. “The Effect of Price, Brand Name, and Store Name on Buyers' Perceptions of Product Quality: An Integrative Review.” Journal of Marketing Research 26, no. 3 (1989): 351-357. DOI 10.1177/002224378902600309.

Reinartz, Werner J., and V. Kumar. “On the Profitability of Long-Life Customers in a Noncontractual Setting: An Empirical Investigation and Implications for Marketing.” Journal of Marketing 64, no. 4 (2000): 17-35. DOI 10.1509/jmkg.64.4.17.18077.

Scheibehenne, Benjamin, Rainer Greifeneder, and Peter M. Todd. “Can There Ever Be Too Many Options? A Meta-Analytic Review of Choice Overload.” Journal of Consumer Research 37, no. 3 (2010): 409-425. DOI 10.1086/651235.

Smith, Wendell R. “Product Differentiation and Market Segmentation as Alternative Marketing Strategies.” Journal of Marketing 21, no. 1 (1956): 3-8. DOI 10.1177/002224295602100102.

Zeithaml, Valarie A. “Consumer Perceptions of Price, Quality, and Value: A Means-End Model and Synthesis of Evidence.” Journal of Marketing 52, no. 3 (1988): 2-22. DOI 10.1177/002224298805200302.

Books and extended reports

Binet, Les, and Peter Field. The Long and the Short of It: Balancing Short and Long-Term Marketing Strategies. London: Institute of Practitioners in Advertising, 2013.

Bucklin, Louis P. A Theory of Distribution Channel Structure. Berkeley: Institute of Business and Economic Research, University of California, 1966.

Coughlan, Anne T., Erin Anderson, Louis W. Stern, and Adel I. El-Ansary. Marketing Channels. 7th ed. Upper Saddle River, NJ: Prentice Hall, 2006.

Dunford, April. Obviously Awesome: How to Nail Product Positioning so Customers Get It, Buy It, Love It. Toronto: April Dunford, 2019.

Kotler, Philip, Kevin Lane Keller, and Alexander Chernev. Marketing Management. 16th ed. Harlow: Pearson, 2022.

McCarthy, E. Jerome. Basic Marketing: A Managerial Approach. Homewood, IL: Richard D. Irwin, 1960.

Nagle, Thomas T., John E. Hogan, and Joseph Zale. The Strategy and Tactics of Pricing: A Guide to Growing More Profitably. 5th ed., New International Edition. Abingdon: Routledge, 2016.

Owen, David. Copies in Seconds: Chester Carlson and the Birth of the Xerox Machine. New York: Simon & Schuster, 2004.