Reference
Frameworks: Playing to Win and Porter's Strategic Coherence
Part of the Strategic Choice skill · loaded on demand from SKILL.md
Contents
- Why these two frameworks, and how they divide the work
- Playing to Win (Lafley & Martin, 2013): the strategy cascade (2.1 writing each choice to exclude something; 2.2 nested cascades; 2.3 two ways to win)
- The strategy logic flow: the analyses that inform the cascade (3.1 defining strategic segments; 3.2 channel and end customer)
- Reverse engineering with "what would have to be true" (Lafley, Martin, Rivkin & Siggelkow, HBR 2012) (4.1 under deadline; 4.2 what the sceptic is for)
- Porter, "What Is Strategy?" (HBR 1996): positioning, trade-offs and fit
- The activity-system map as a test instrument (6.1 constructing the map; 6.2 what the map is not)
- How the two frameworks are combined in this skill
- Common misreadings
- Sources
1. Why these two frameworks, and how they divide the work
Strategy work fails in two characteristic ways. The first is that no choice is made: the plan lists initiatives in every segment, offends no one, and excludes nothing. The second is that a choice is made, but by advocacy: the most senior or most persuasive person wins a debate conducted in pros and cons, and the organisation inherits a strategy whose logic nobody can reconstruct. Playing to Win is designed against both. The cascade forces five explicit choices, each of which excludes something; the reverse-engineering method replaces debate with conditions and tests.
Porter's 1996 article addresses a third failure, which Playing to Win does not fully guard against: a cascade can be completed with answers that are true but generic. "Where to play: mid-market customers. How to win: superior service." These pass the form of the cascade and fail as strategy, because nothing in them forces a trade-off and no rival is prevented from copying them. Porter supplies the test: is the position based on distinct activities, does it require trade-offs, and do the activities fit together so that the whole is harder to imitate than the parts?
The division of labour in this skill is therefore: Playing to Win for generating and choosing; Porter for checking that what was chosen is a strategy at all.
2. Playing to Win (Lafley & Martin, 2013): the strategy cascade
A.G. Lafley, CEO of Procter & Gamble from 2000 to 2009 (and again from 2013 to 2015), and Roger Martin, dean of the Rotman School of Management from 1998 to 2013 and Lafley's strategy adviser, published Playing to Win: How Strategy Really Works (Harvard Business Review Press, 2013), based on their work together at P&G from 2000 onward. The central claim is that strategy is a set of integrated choices that position a firm to win, and that the set can be written as five questions in a cascade:
-
What is our winning aspiration? The purpose of the enterprise, expressed as winning rather than participating. Lafley and Martin insist on "winning" because an aspiration to "be a player" or "serve customers well" does not force the later choices. Winning is defined relative to competitors and in the customer's terms, and it should be specific enough to be measured. The aspiration is not the strategy; it is what the strategy must deliver.
-
Where will we play? The playing field: which geographies, product categories, customer segments, channels, and stages of the value chain. The book treats where-to-play as a choice among fields, not a description of the market, and stresses that the choice must be narrow enough to be actionable. Choosing everywhere is choosing nowhere.
-
How will we win? The source of advantage on the chosen field: what the firm will do that causes the chosen customer to choose it. The authors adopt Porter's two generic routes, low cost and differentiation, and hold that a firm must pick one as the primary logic. How-to-win must be answered together with where-to-play, because a how-to-win that works on one field may fail on another, and a where-to-play is only attractive if there is a way to win on it.
-
What capabilities must be in place? The activities the organisation must perform at a distinctive level for how-to-win to hold. The book presents these as an "activity system," borrowing directly from Porter, and argues that capabilities should be few (P&G's were consumer understanding, brand building, innovation, go-to-market, and scale) and reinforcing.
-
What management systems are required? The processes, structures, measures, and norms that build and sustain the capabilities and that keep the choices alive as conditions change. The book's point is that strategy decays without systems: a capability nobody measures or reviews stops being distinctive within a few years.
The mechanism that makes the cascade work is mutual constraint. The arrows run both ways: aspiration shapes where-to-play, but a where-to-play that cannot be won at reshapes the aspiration; how-to-win demands capabilities, but capabilities that cannot be built in time reshape how-to-win. A cascade that reads coherently top-down but not bottom-up (the capabilities do not actually deliver the advantage) is the most common way it is filled in badly. Lafley and Martin also stress that the cascade applies at every level of an organisation: a business unit, a brand, or a function has its own nested cascade, and each must be consistent with the one above.
The failure the cascade prevents is the strategy document that is a mission statement plus a budget. Such documents answer the aspiration question with values and the where-to-play question with the current footprint, then jump to initiatives. Nothing is excluded, so nothing is chosen.
2.1 Writing each choice so that it excludes something
The practical test for every cascade answer is whether it excludes something a reasonable competitor might choose. Some guidance per choice, drawn from the book's treatment and from how the answers fail in practice:
- Winning aspiration. Write it as winning against someone, in someone's eyes, by some date, measured somehow. "Be the quality system of record for mid-market life-sciences manufacturers in North America and Europe by fiscal 2028, measured by customer count" excludes being the biggest by ARR and excludes enterprise accounts and other industries. "Be our customers' trusted partner" excludes nothing. The book's P&G examples are of this kind: for the Olay brand, winning meant beating the prestige skin-care brands on the mass shelf, not merely growing.
- Where to play. Name the dimensions (customers, geographies, categories, channels, value-chain stages) and for each say what is in and what is out. The "out" list is the strategy. A where-to-play that lists only what is in has not been chosen; it has been described.
- How to win. State the generic route (low cost or differentiation) and then the specific mechanism: which customer, choosing us over which alternative, because of which activity they cannot get elsewhere. Lafley and Martin follow Porter in holding that trying to be both the low-cost player and the differentiated player usually yields neither; the exceptions exist but must be argued, not assumed.
- Capabilities. Few, named as activities performed at a distinctive level, and each traceable to the how-to-win. Mark which exist and which must be built, with cost and time. A capability list longer than six is usually a list of functions.
- Management systems. Name the measures, reviews, incentives and processes, and for each say which capability it sustains. Include the review that checks whether the strategy's conditions still hold; without it the strategy decays silently.
2.2 Nested cascades
The book is explicit that the cascade applies at every level: corporation, business unit, brand, function. Each lower-level cascade must be consistent with the one above, and the where-to-play of the corporation constrains the aspiration of the unit. In consulting work this matters when the client is a division: the division's options are bounded by the parent's choices, and a condition of any option may be "the parent will fund and permit this." Write that condition down; it is often the barrier.
2.3 Two ways to win
Playing to Win adopts Porter's generic advantages from Competitive Strategy (1980): a firm wins either by having lower costs than rivals for an equivalent offer, or by offering something customers value enough to pay a premium for. The authors' own emphasis is that the choice must be made and lived: a cost leader's capabilities and systems (standardisation, scale, relentless cost measurement) differ from a differentiator's (customer insight, brand, service design), and an organisation that tries to build both sets builds neither to a distinctive level. This is the same argument Porter makes about trade-offs, from the inside of the firm rather than from the market.
3. The strategy logic flow: the analyses that inform the cascade
Chapter 7 of Playing to Win describes the "strategy logic flow," the set of analyses that inform where-to-play and how-to-win. It has four parts, examined iteratively rather than in strict sequence:
- Industry analysis. What strategic segments exist, and what is the structure of each (how attractive it is, and why). Segments are defined by differences in what wins, not by convenience of reporting. The structural analysis is essentially Porter's Five Forces applied per segment.
- Customer value analysis. What channel customers value, and what end customers value, examined separately because they often differ. The book's emphasis is on going deep with real customers rather than accepting the aggregate market research.
- Analysis of relative position. How the firm's capabilities and its cost position compare with rivals in the segment. This is the honest reckoning of whether the firm can actually deliver a given how-to-win.
- Competitor analysis. How rivals are likely to react to each possible choice, and whether that reaction erodes the advantage.
In this skill the four headings are used as the taxonomy for WWHTBT conditions. Every option's conditions are sorted under industry, customer value, relative position, and competitor reaction. The purpose is completeness: teams reliably write customer and economics conditions and reliably forget competitor reaction, which is where the most expensive surprises come from.
3.1 Defining strategic segments
The industry analysis depends on segmenting correctly, and the book's guidance is that a strategic segment is a part of the market in which the basis of winning differs. Two customers are in the same segment if the same activity system would win both; they are in different segments if winning one requires activities that would be over- or under-designed for the other. This is a different cut from the reporting segments most clients use (region, product line, account size), and it is why the fact base from market-competitive-intelligence should be re-cut before options are generated. In the Sablecrest case the reporting segments were modules and sales regions; the strategic segments turned out to be companies by regulatory exposure and size, because a regulated manufacturer with 200 to 2,000 employees needs a validated system and a certified implementation and a 60-person company or a non-regulated one does not, whichever modules it buys.
Attractiveness of a segment is then assessed structurally (Five Forces per segment) and in terms of the client's relative position in it. A segment can be attractive in general and unwinnable for this client, or unattractive in general and winnable because rivals have left it. Both facts belong in the conditions.
3.2 Customer value: channel and end customer separately
The book insists on examining the channel customer and the end customer separately because their interests diverge and a strategy that satisfies one at the other's expense fails. For a software vendor like Sablecrest the "channel" is the ERP vendors and implementation partners whose referrals and connector certifications it depends on (whether they will list it, certify it and refer it) and the end customer is the quality leader at the manufacturer. For a consumer brand it is the retailer and the shopper. Conditions under this heading should therefore come in pairs: what the end customer must value, and what the channel must gain.
4. Reverse engineering with "what would have to be true" (Lafley, Martin, Rivkin & Siggelkow, HBR 2012)
The method for choosing among options is set out in "Bringing Science to the Art of Strategy," by A.G. Lafley, Roger Martin, Jan Rivkin and Nicolaj Siggelkow, Harvard Business Review, September 2012, and restated as chapter 8 of Playing to Win. Martin has described its origin as an engagement with Inmet Mining in 1994, where a management team was deadlocked between advocates and sceptics of a possible investment until the question was changed from "is this a good idea?" to "what would have to be true for this to be a good idea?" (Martin, "The Evolution of the Strategic Choice Structuring Process," 2025). His 2022 piece "What Would Have to be True?" covers the question's use in reviewing a strategy mid-stream, and his current version of the process compresses the seven steps to five (define the problem, generate possibilities, design tests, assess, choose); the seven-step form below is the one in the HBR article and the book, and it is the one this skill runs because it keeps barrier identification as an explicit step.
The published procedure has seven steps:
-
Frame a choice. Convert the issue into at least two mutually exclusive options. The status quo is always one of them, described honestly, including where it leads if conditions continue. Framing as a choice matters because it makes clear that something will be given up, and because an issue without options attracts analysis without decision.
-
Generate possibilities. Widen the set. A possibility is a "happy story": a description of where the firm would compete and how it would win, with internal logic, told without yet asking whether it is true. The authors ask for a range, including options the group finds uncomfortable, and they warn against evaluating during generation, because early evaluation eliminates unfamiliar options before their logic is understood.
-
Specify conditions. For each possibility, describe what would have to be true for it to be a great choice. The conditions are written by the whole group, including the sceptics, and the question is deliberately not "is this option good?" but "what would need to hold for you to be enthusiastic about it?" The HBR article organises conditions under headings matching the strategy logic flow: industry (segments, structure), customer value (channel, end customers), relative position (capabilities, costs), and competitor reaction. A condition is a statement that could be checked, not an opinion.
-
Identify barriers. Of the conditions for each option, which ones does the group most doubt? These are the barrier conditions. The article notes that the group's confidence often diverges, and that divergence is informative: the person least confident in a condition is the one whose view the test must be designed to change.
-
Design tests. For each barrier condition, design a test that the sceptic agrees would be valid. The pass mark is set in advance. Tests can be analyses, interviews, pilots, or experiments; the criterion is that the sceptic accepts the design, not that the test is comprehensive.
-
Conduct the tests. Run them in reverse order of confidence: the condition the group trusts least is tested first. The authors call this the "lazy man's approach," because if the shakiest condition fails, the option is eliminated without spending anything on the others. Stop testing an option as soon as one of its barrier conditions fails.
-
Make the choice. With the results in hand, the choice usually becomes clear: pick the option whose conditions have survived. The article observes that by this point the group has done the work together, and the choice is a matter of reviewing results rather than arguing.
The reasoning behind the method is worth restating, because it is what lets the model apply it with judgment. Advocacy debates fail because each side is arguing about the whole option, and the whole option is too large to be settled by evidence in the time available. Conditions are small enough to be checked. Moreover, asking the sceptic what would have to be true converts an opponent into a co-author: their objection becomes a condition, their doubt becomes a test, and if the test passes they have already agreed what it means. This is why the method must not be shortcut by having the advocate write the conditions alone. A condition list without the sceptic's objection in it is a pros list with a new heading.
The difference from a pros-and-cons analysis is therefore structural, not cosmetic: pros and cons are evaluations offered by people; conditions are facts about the world that either hold or do not, and the evaluation is derived from them.
4.1 Running the seven steps under deadline
The published method assumes a series of workshops over weeks. Consulting engagements often have days. The steps that cannot be compressed are 3 (conditions written with the sceptic) and 5 (pass marks set before results), because compressing them reintroduces advocacy. The steps that compress well:
- Step 1 can be done in an hour with the decision-maker if the problem statement from
problem-framingexists. - Step 2 can be done in a half-day workshop using the three positioning bases as prompts; the facilitator writes the status quo card so it is not left to the person who least wants it examined.
- Step 4 (barriers) is a 20-minute private vote followed by a 30-minute comparison. The comparison, not the vote, is where the information is.
- Step 6 can often be run on existing evidence for most conditions; the one or two conditions that need new evidence are the barriers, and a pilot or a set of 30 interviews can be done in six to eight weeks. Where even that is unavailable, the condition is recorded as an assumption with a signpost, and the choice is made explicitly conditional. A conditional choice with a named trigger for revisiting is honest; a confident choice resting on an untested barrier is not.
4.2 What the sceptic is for
The method's reliance on the sceptic is easily lost in application, so it is worth being explicit about the mechanism. An advocate writing conditions writes the ones the option satisfies. A sceptic writing conditions writes the ones it may not. The union is the real list. When the sceptic designs the test and sets the pass mark, three things follow: the test is one they will believe; a pass commits them, because they chose the threshold; and a fail retires the option without the advocate being able to claim the test was unfair. If no sceptic exists for an option, appoint one, and give them the fact base; an option with no sceptic is usually one nobody has examined.
5. Porter, "What Is Strategy?" (HBR 1996): positioning, trade-offs and fit
Michael Porter's "What Is Strategy?" appeared in Harvard Business Review, November-December 1996. It was written against the background of a decade in which management tools (TQM, benchmarking, reengineering) had improved operational effectiveness across whole industries without improving the profitability of most firms in them.
Operational effectiveness is not strategy. Operational effectiveness means performing similar activities better than rivals. It is necessary, but it converges: best practices diffuse, the productivity frontier moves outward for everyone, and competition on operational effectiveness alone becomes mutually destructive. Strategy is performing different activities from rivals, or similar activities in different ways. The distinction matters for this skill because a great many "strategies" presented to boards are operational-effectiveness programmes: cost-out, digital parity, service-level improvement. They are worth doing, but they are not a choice, and they will not produce a durable difference.
Three bases of positioning. Porter identifies three sources of a distinctive position, which are not mutually exclusive but are useful as generators:
- Variety-based: producing a subset of an industry's products or services, chosen because a distinctive set of activities can produce them best. Jiffy Lube (only lubrication services, faster and cheaper than full-service garages); Vanguard (index and low-turnover funds, no bets on interest rates or sectors).
- Needs-based: serving most or all of the needs of a particular group of customers, where the group has different needs and a tailored set of activities serves them best. IKEA (young, price-sensitive furniture buyers who will trade service for price); Bessemer Trust (families with 5 million dollars or more, capital preservation through personal service) versus Citibank private banking (250,000 dollars or more, convenient credit).
- Access-based: serving customers who are reachable in different ways, because of geography or scale, even if their needs are similar to others'. Carmike Cinemas (towns under 200,000 people, standardised low-cost theatres with a lean central overhead).
Trade-offs make positions sustainable. A position is only defensible if occupying it means giving something up. Porter gives three reasons trade-offs arise:
- Inconsistencies in image or reputation: a brand known for one thing cannot credibly deliver its opposite (Ivory could not become Neutrogena).
- Activities themselves: different positions require different product configurations, equipment, skills, and systems; an activity designed for one position is over- or under-designed for another (IKEA's self-service floor cannot also deliver high-touch selling).
- Limits on internal coordination and control: an organisation that has chosen clearly can let employees make everyday decisions; one that tries to be everything to everyone produces confusion in the trenches. The case is Continental Lite: Continental tried to match Southwest on selected routes while keeping its full-service position elsewhere, and could not do both, because travel-agent commissions, frequent-flier programmes, hub operations, and baggage transfers were all designed for the full-service position. Porter calls this straddling, and it is the most common form of failed strategy in practice: adding a rival's benefits onto an existing activity system.
Fit. Strategic positions are built from systems of activities, not individual ones, and fit among activities is what makes a position hard to copy. Porter describes three orders of fit:
- First order, simple consistency: each activity is aligned with the overall strategy. Vanguard's low turnover, direct distribution, limited advertising, and cost-linked bonuses all point at low cost.
- Second order, reinforcement: activities make each other more effective. Neutrogena's placement in upscale hotels reinforces its dermatologist-recommended positioning and lowers the cost of its drugstore marketing.
- Third order, optimisation of effort: coordination across activities eliminates redundancy and wasted effort. The Gap restocks basics daily from three warehouses, which lets stores hold minimal inventory and supports short product cycles. The sustainability argument is arithmetic: a rival who can match each of four interlocked activities with 90 percent probability matches the system with 66 percent probability, and the odds fall fast as activities are added.
The growth trap. Porter observes that the desire to grow is the most corrosive pressure on strategy: firms extend lines, add features, imitate rivals' services, and acquire, each step blurring the position. Maytag's expansion from reliable laundry into a full appliance range grew revenue from 684 million dollars (1985) to 3.4 billion (1994) while return on sales fell from 8 to 12 percent to under 1 percent. Neutrogena diluted a dermatologist-recommended position by extending into mass categories and mass channels. His prescription is to grow by deepening the position (more penetration of the chosen customers, extensions that use the existing activity system) or by geographic expansion, and, where a genuinely different position is wanted, to house it in a separate unit with its own brand and activities.
Continuity and leadership. Porter closes by arguing that strategic positions should be held for years, not planning cycles, because fit among activities takes time to build and is destroyed quickly by hedging. He assigns the job of holding the line to general management: the leader's role includes saying no to the extensions, accounts and features that would blur the position, and explaining the trade-offs to the organisation so that day-to-day decisions are made consistently. In the skill this appears as the trade-off register, which is written so that the standing answer to each request is known before it arrives, and as the growth-trap signpost, which forces any broadening proposal through the full method rather than through a budget discussion.
Reading the positioning bases in a fact base. Each base leaves a signature in the numbers. A variety-based opportunity shows up as a category cluster with structurally higher gross margin and a different buying process (technical, specification-driven) from the rest of the line. A needs-based opportunity shows up as a customer group with higher retention, higher wallet share where served well, and a different cost-to-serve profile, which is what Sablecrest's regulated life-sciences segment shows (118 percent net revenue retention against 101 percent elsewhere). An access-based opportunity shows up as geography or size classes where the larger rivals' deal-size or account thresholds leave demand unserved, and where the client's cost of reaching them is lower than a rival's would be. When generating options, look for these signatures before inventing stories.
6. The activity-system map as a test instrument
Porter's article introduces activity-system maps: diagrams in which a few higher-order strategic themes (for Southwest: limited passenger service, frequent reliable departures, low ticket prices, lean ground crews, high aircraft utilisation, short-haul point-to-point routes between mid-size cities and secondary airports) are shown as large nodes, and the tailored activities that deliver them (no meals, no seat assignments, 15-minute gate turnarounds, standardised 737 fleet, secondary airports, flexible union contracts) as smaller nodes, with lines showing which activities support which themes and each other.
In this skill the map is drawn in table form (theme; activities that deliver it; other activities that reinforce it; trade-off it requires) because tables survive copying into documents and can be checked line by line. The tests applied to the map are Porter's three orders of fit, plus a trade-off check and an interlock check:
- Consistency: does any activity work against how-to-win? If so, the cascade is not describing the real strategy.
- Reinforcement: name at least three pairs of activities where one makes the other cheaper or better.
- Optimisation: name at least three activities the position allows the firm to stop doing.
- Trade-offs: list what the firm will not do, and confirm each item costs something real today.
- Interlock: name the activity a rival would find hardest to copy and the activities that make it hard.
A cascade whose activity map fails these tests is a plan, and possibly a good one, but it is not a strategic position, and its advantage will be competed away.
6.1 Constructing the map
- Start from how-to-win and ask what must be true operationally for it to hold. Each answer is a candidate theme. Three to five themes; more means the position is not yet clear.
- For each theme, list the activities that deliver it. Write each as a practice with a specific design choice in it ("one certified consultant per 12 accounts", "a validation package produced for every release on a fixed quarterly cadence"), because the design choice is what a rival must copy. If an activity could be performed by any competitor in the same way, it belongs in the operational-effectiveness column, not the map.
- For each activity, ask which other activities it depends on or makes cheaper. Record the pairs. These are the reinforcement links, and they are the map's evidence of fit.
- For each theme, ask what the firm must stop doing or refuse to do for the theme to hold. Record it as a trade-off with its cost.
- Apply the five tests. Where a test fails, the fix is usually in the cascade, not in the map: an activity that contradicts how-to-win means the cascade is describing a strategy the firm is not actually pursuing.
6.2 What the map is not
It is not an organisation chart, a process map, or a value chain. An organisation chart shows who reports to whom; a process map shows sequence; a value chain shows the generic stages of creating value. The activity map shows which distinctive practices depend on which others. A map whose nodes are "Sales", "Operations" and "IT" has been drawn from the org chart and will pass none of the fit tests, because functions do not reinforce each other in any specific way.
7. How the two frameworks are combined in this skill
| Step in the skill | Playing to Win contributes | Porter contributes |
|---|---|---|
| Frame the choice | The insistence on mutually exclusive options including the status quo | The reminder that operational-effectiveness programmes are not options |
| Generate possibilities | The happy-story format; generate before evaluating | The three positioning bases as generators; straddles are excluded |
| Specify conditions | The WWHTBT question and the four logic-flow headings | Trade-off and imitability conditions under relative position and competitor reaction |
| Identify barriers, design and run tests | The sceptic-designed test, pre-set pass marks, least-confident first | |
| Make the choice and write the cascade | The five questions and bottom-up coherence check | The requirement that each answer excludes something |
| Activity system and fit | Capabilities as an activity system | The map, three orders of fit, trade-offs, interlock |
| Record and signposts | Management systems that keep conditions under review | The growth-trap warning as a standing signpost |
8. Common misreadings
- Treating the cascade as a form. Filling in five boxes is not the method. The method is the constraint each box places on the others, checked in both directions.
- Winning aspiration as values. "To delight customers" is not an aspiration to win. Ask: against whom, in whose eyes, by when, measured how?
- Where-to-play as market description. The current footprint is not a choice. Where-to-play must name what is out.
- Confusing capabilities with functions. "Sales" is a function. "A maintained validation package shipped with every release to 560 regulated customers" is a capability.
- WWHTBT written by the advocate. The conditions then confirm the option. The sceptic's objection must be in the list, in their words.
- Testing everything. Comprehensive analysis of all conditions for all options is what the method exists to avoid. Test the least-confident condition first and stop when an option dies.
- Reading Porter as "focus is always right." Porter does not say narrow is better; he says a position must require trade-offs. A broad position with a coherent activity system (Vanguard is broad in customers, narrow in product philosophy) is a valid position.
- Mistaking the positioning bases for a menu of three. Real positions often combine bases (IKEA is needs-based and variety-based). The bases are generators for options, not categories to file the final answer under.
- Activity-system map as organisation chart. If the nodes are departments, redraw with activities.
9. Sources
- Lafley, A.G. and Martin, R.L. (2013). Playing to Win: How Strategy Really Works. Harvard Business Review Press. Chapters 1 to 6 for the cascade; chapter 7 for the strategy logic flow; chapter 8 for reverse engineering.
- Lafley, A.G., Martin, R.L., Rivkin, J.W. and Siggelkow, N. (2012). "Bringing Science to the Art of Strategy." Harvard Business Review, September 2012.
- Martin, R.L. (2022). "What Would Have to be True? The Most Valuable Question in Strategy." Playing to Win/Practitioner Insights series, Medium and Substack, August 2022.
- Martin, R.L. (2025). "The Evolution of the Strategic Choice Structuring Process." Playing to Win/Practitioner Insights series, Medium and Substack, June 2025. Source for the Inmet Mining origin (1994) and the current five-step form.
- Porter, M.E. (1996). "What Is Strategy?" Harvard Business Review, November-December 1996.
- Porter, M.E. (1980). Competitive Strategy. Free Press. For the two generic advantages (low cost, differentiation) that Playing to Win adopts.