Playing to Win
The strategy choice cascade: aspiration, where to play, how to win, capabilities, systems
Playing to Win is a way to set strategy as five linked choices instead of a long plan. You decide what winning means, where you will compete, how you will win there, which capabilities you must have, and which management systems will support them. Where to play and how to win are chosen together, as a matched pair, and the last two choices test whether that pair is real.
TL;DR
- What it is
- Five linked choices that make up a strategy, from what winning means to the systems that support it. Set out by A.G. Lafley and Roger Martin in 2013.
- Best for
- Choosing between real directions, when a long plan has stopped saying what the company will not do.
- First cycle
- One working day to frame the options and what would have to be true. A few weeks more to test the doubtful conditions.
- You leave with
- One matched choice of where to play and how to win, and the capabilities and systems it needs.
- Winning aspiration
- Where to play
- How to win
- Must-have capabilities
- Management systems
Most strategic plans are a goal, a long list of initiatives and a budget. Playing to Win replaces that with a short set of choices, and the choices include what you will not do. Martin argues that a strategy you can write on one page, in plain words, is more useful than a thick plan4.
The part most teams miss is how the choice is made. Instead of arguing about which option is right, the team lists what would have to be true for each option to work, then tests only the conditions it doubts most3. Disagreement turns into a short list of tests.
The original Playing to Win, in brief
From "Playing to Win: How Strategy Really Works" (Harvard Business Review Press, 2013) by A.G. Lafley, former chief executive of Procter & Gamble, and Roger L. Martin, former dean of the Rotman School of Management. An expanded edition followed in 2025.
A.G. Lafley, chief executive of Procter & Gamble from 2000 to 2009, and Roger L. Martin, then dean of the Rotman School of Management at the University of Toronto, published Playing to Win: How Strategy Really Works with Harvard Business Review Press in 20131. It builds the whole idea around five questions: what is your winning aspiration, where will you play, how will you win, what capabilities must be in place, and what management systems support those choices1. An expanded edition with two added HBR articles came out in 20251.
Both halves of the method appeared in Harvard Business Review first. In 2010 Martin described strategy as a choice cascade: choices flow down through the company like water through rapids, and each level chooses within the choices made above it2. In 2012 Lafley and Martin, with Jan Rivkin and Nicolaj Siggelkow, set out how to pick between options by asking what would have to be true for each one to work, and then testing only the conditions the team doubts most3.
The core claim is that strategy is a small set of choices, not a plan. Martin argued in 2014 that two of them decide success: which customers to serve, and why those customers would pick you over the alternatives4. The other three check whether that pair is real. If you cannot build the capabilities and systems it needs, Martin wrote, "it is a fantasy, not a strategy"5.
Common misreadings
- "Fill in the boxes from top to bottom." Martin wrote in 2017 that teams who settled where to play first, then turned to how to win, got stuck. The two only make sense as a matched pair, and the choices have to reinforce each other in both directions5.
- "Capabilities and systems are execution, not strategy." Martin rejects that split. He treats them as part of the strategy and as its reality check5, and he argued in 2010 that the line between strategy and execution does real damage2.
- "The cascade is for the top team." In the original, every level makes choices, down to the person serving a customer. Each works within the choices above, and what they learn flows back up2.
- "A strategic plan is a strategy." Martin described the typical plan as a lofty goal, a long list of initiatives and a set of financial projections. It rarely says what the company will not do or which assumptions it rests on4.
- "What would have to be true means what do you believe." It is the reverse. The team lists the conditions each option needs, and nobody argues yet about whether they hold. Evidence comes later, in tests3.
The Playing to Win template
Winning aspiration
What winning means for you, and for whom. Specific enough that you could tell if you had not won.
Where to play
The customers, markets, products and channels you will compete in, and the ones you will leave alone.
How to win
Why the customers you chose will pick you over the alternatives, there. Chosen together with where to play.
Must-have capabilities
The few things you must be good at for that choice to work. What you have, and what you must build.
Management systems
The processes, measures and routines that build and keep those capabilities.
Playing to Win examples
A 30-person accounting firm
Fee pressure on compliance work. The choice is to move into advice for a narrower set of clients.
A 150-person manufacturer
Selling pumps on price against larger rivals. The choice is to win on service in a narrower market.
A 60-person software company
Selling to everyone and winning no segment. The choice is one industry, served deeply.
When Playing to Win fits, where it struggles, and what it does not answer
Use it when
- The leadership team approves a longer list of initiatives every year but cannot say what it will stop doing.
- You face two or more real directions, such as a new customer group, a new market or a new position on price.
- Growth has stalled and the way you used to win no longer sets you apart.
- Debates about direction go in circles. Asking what would have to be true turns opinions into conditions you can test3.
- You are about to set yearly goals or budgets and want the choices settled first, so the goals serve them.
It struggles when
- It assumes you can choose in advance. Henry Mintzberg argued in 1994 that many good strategies emerge from learning across the organization rather than being planned at the top6. Martin's answer is to write the logic down and check it against events4. A young company still searching for its market may learn faster by trying things.
- It has no diagnosis step. The cascade starts from the aspiration. Richard Rumelt argued in 2011 that good strategy starts with a clear diagnosis of the challenge7. Without a shared view of what is going on, where-to-play debates drift.
- It was built from one company. The main case is P&G, a very large maker of branded consumer goods1. Small firms, service businesses and public bodies have to translate the examples.
- It names management systems but does not design them. The fifth question asks which systems support the choices1. It does not tell you what to measure, who owns what, or how often to review.
- Good tests take time. The method asks the team to test its most doubtful conditions, which can mean customer research or a pilot3. A one-day retreat can frame the choice but rarely finish it.
What it does not answer
- How to run the work week to week once the choice is made.
- What to measure, and who owns each piece.
- Whether your structure, people and culture fit the choice.
- What is changing in your market that nobody has noticed yet.
Pair it with
- OKR Turn the capabilities and systems you must build into quarterly objectives with measurable results.
- McKinsey 7S Check whether your structure, systems, style and people can carry the choice before you commit to it.
- Blue Ocean Strategy A way to generate a where-to-play option outside the market as it is drawn today.
How to implement Playing to Win
- Turn the issue into a choice. Write at least two options that would each resolve it, then widen to three to five. Always include carrying on as you are3.
- Write each option as a matched pair. For every option, say where you would play and how you would win there, together. A where to play on its own cannot be judged5.
- List what would have to be true. For each option, list the conditions it needs about customers, competitors, costs and your capabilities. Do not argue yet about whether they hold3.
- Test the doubtful conditions. Rank the conditions by how much the team doubts them. Let the biggest skeptic design the test, and test the least likely condition first3.
- Choose, then fill in the rest of the cascade. Pick the option that survives the tests. Name the capabilities and systems it needs, and ask each level below to set its own choices within it2.
Adopting Playing to Win: the first cycle and the rhythm
First cycle. Frame the decision as a choice between at least two options, then widen it to three to five, and always include carrying on as you are3. For each, list what would have to be true. A working day gets you that far. Testing the two or three conditions the team doubts most takes weeks, depending on whether a test is a few customer calls or a pilot3.
Rhythm. The framework sets no cadence. Write down the conditions your choice depends on and compare them with what actually happens4. A quarterly review and a yearly retreat are natural points to check. Reopen the choice when a condition stops holding, not on a calendar.
How long until it runs without effort. The first cascade is usually rough. It starts to work as a shared language once the leaders below the top team write their own choices inside yours, and tell you what they learn2.
What derails it. Settling where to play before how to win5. A how to win that is really a list of initiatives4. Leaving the status quo off the list of options, so it never faces the same test3. Letting the most senior person lead the discussion; the 2012 article advises against it3.
How to set Playing to Win at your leadership retreat
The decision the session has to produce: Which matched pair of where to play and how to win the company will pursue, what would have to be true for it, and which conditions to test first.
A half-day outline
- Restate the issue and why a choice is needed now (20 minutes).
- Everyone sketches two or three options alone, each as where to play and how to win (30 minutes).
- Share and merge them into three to five options, including carrying on as you are (45 minutes).
- Break.
- For each option, list what would have to be true. No arguing about whether it is true yet (60 minutes).
- Rank the conditions the team doubts most and agree a test for each (40 minutes).
- Give every test an owner and a date, and book the session where you will choose (15 minutes).
Ask the team beforehand
- Which customers would you stop serving if you had to choose?
- Why do your best customers pick you today, and would they say the same?
- What is the company planning to do this year that no strategy choice asked for?
Who should be in the room
The chief executive and the leaders who own customers, product and operations. Keep the group to eight or ten; the 2012 HBR article advises that larger groups tend to hold back.
Make a full retreat agenda with the free agenda maker.
Common Playing to Win mistakes
- Choosing where to play on its own. A market is only attractive if you can win there. Teams that settle where to play first often get stuck at how to win5.
- A how to win that is a list of projects. Launches and expansions are initiatives. How to win says why a customer would choose you4.
- Leaving the current path off the list. Carrying on as you are is a choice with its own risks. Put it through the same test as the others3.
- Arguing about what is true too early. Debate whether conditions hold before listing them, and the skeptic and the champion just dig in. List first, test second3.
- Treating capabilities and systems as someone else's job. If you cannot have or build them, the choice does not work. They belong in the strategy5.
Does Playing to Win work? The evidence
The book's evidence is P&G. Harvard Business Review Press describes Lafley, working with Martin, as having doubled P&G's sales, quadrupled its profits and added more than $100 billion in market value in ten years1. That is the authors' own account of a company one led and the other advised. It has no comparison group, and much else changed at P&G in those years, so it cannot show that the framework caused the results.
The same holds for the best-known worked case. In their 2012 HBR article, the authors describe how P&G tested price points for a repositioned Olay, launched at $18.99, and saw the brand pass $2.5 billion in yearly sales in under a decade3. It shows how the method runs. It is told by the people who made the decision.
The research for this page found no peer-reviewed study that tests the choice cascade, or the "what would have to be true" method, against companies that did not use it. The nearest research looks at formal strategic planning. A 2019 meta-analysis in Public Administration Review pooled 87 correlations from 31 studies and found a positive, moderate effect of strategic planning on performance8. That tests planning, which Martin argues is a different activity from strategy4, so it neither supports nor undercuts Playing to Win.
Read Playing to Win as a coherent, widely used way to structure strategic choices, written by experienced practitioners. It is not a tested model of cause and effect. Mintzberg's work on emergent strategy is the strongest published counterweight to any choose-first method6.
Playing to Win compared
| Playing to Win | Blue Ocean Strategy | OKR | |
|---|---|---|---|
| The question | Where will you compete and how will you win there? | How can you make the competition irrelevant? | What will you achieve this quarter, and how will you know? |
| What you get | Five linked choices and the conditions they rest on | A new offer and market space to aim at | Objectives with measurable key results |
| Horizon | Several years, reopened when a condition breaks | Several years | Usually a quarter |
| What it leaves out | Measures and a weekly rhythm | How to run the work | Which strategy to pursue |
After the retreat: Playing to Win in Throughline
What it holds. The capabilities and systems the choice needs become priorities, each with an owner, a date and a measure, and so do the tests you agreed. Owners check in from an email in one click; a missed check-in counts as off track and their manager is told.
What it doesn’t. Throughline does not store the five choices as a cascade, or the conditions you said would have to be true. Keep those on one page of your own and check them against what happens at each review.
Playing to Win glossary
- Winning aspiration
- The purpose of the strategy: what winning would look like for the company.
- Where to play
- The choice of customers, markets, products, channels and stages of the business you will compete in, and by implication the ones you will not.
- How to win
- Why the customers you chose will pick you over the alternatives, in that place.
- Capabilities
- The few activities the company has to be good at to make its where to play and how to win work.
- Management systems
- The processes, measures and routines that build and keep those capabilities.
- Strategy choice cascade
- The five choices together, repeated at each level of the company, each level working within the one above2.
- What would have to be true
- The question used to compare options: the conditions each would need, listed before anyone argues whether they hold3.
- Possibility
- A candidate strategy told as a short story of where you would play and how you would win3.
- Barrier condition
- A condition the team doubts most. It gets tested first, because if it fails the option is out3.
Free tools that help
Playing to Win: frequently asked questions
- What is Playing to Win?
- A strategy framework from the 2013 book by A.G. Lafley and Roger L. Martin. It treats strategy as five linked choices: a winning aspiration, where to play, how to win, the capabilities you must have, and the management systems that support them.
- What are the five questions of Playing to Win?
- What is your winning aspiration? Where will you play? How will you win? What capabilities must be in place? What management systems support those choices? Together they are often called the strategy choice cascade.
- What is the difference between where to play and how to win?
- Where to play is the choice of customers, markets, products and channels. How to win is why those customers will choose you there. Martin argues they must be chosen together: neither is strong or weak on its own.
- What does "what would have to be true" mean?
- It is the question the authors use to choose between options. For each option, the team lists the conditions it would need, without arguing yet about whether they hold. Then it tests the ones it doubts most, starting with the least likely.
- Do you answer the five questions in order?
- No. The diagram runs top to bottom, but the choices have to fit each other. Draft where to play and how to win together, then check them against the aspiration, the capabilities and the systems, and revise.
- How is Playing to Win different from a strategic plan?
- A typical plan lists goals, initiatives and budgets. Playing to Win asks for a few explicit choices, including what you will not do, and for the assumptions behind them to be written down and checked.
- Does Playing to Win work for small companies?
- The questions do. The book's examples come from a very large consumer goods company, so a small firm has to translate them, and the tests will be simpler: a few customer calls instead of a large study.
- Is Playing to Win better than OKRs?
- They do different jobs. Playing to Win chooses the strategy. OKRs turn it into quarterly goals with measurable results. Many teams use Playing to Win to set the choices and OKRs to build the capabilities it calls for.
Sources
- A.G. Lafley and Roger L. Martin, Playing to Win: How Strategy Really Works, Harvard Business Review Press, 2013; expanded edition, 2025. store.hbr.org
- Roger L. Martin, "The Execution Trap," Harvard Business Review, July to August 2010. hbr.org
- A.G. Lafley, Roger L. Martin, Jan W. Rivkin and Nicolaj Siggelkow, "Bringing Science to the Art of Strategy," Harvard Business Review, September 2012. hbr.org
- Roger L. Martin, "The Big Lie of Strategic Planning," Harvard Business Review, January to February 2014. hbr.org
- Roger L. Martin, "Strategic Choices Need to Be Made Simultaneously, Not Sequentially," Harvard Business Review, April 2017. hbr.org
- Henry Mintzberg, "The Fall and Rise of Strategic Planning," Harvard Business Review, January to February 1994. hbr.org
- Richard Rumelt, Good Strategy/Bad Strategy: The Difference and Why It Matters, Crown Business, 2011.
- Bert George, Richard M. Walker and Joost Monster, "Does Strategic Planning Improve Organizational Performance? A Meta-Analysis," Public Administration Review 79(6), 2019. doi.org
Playing to Win and the strategy choice cascade were set out by A.G. Lafley and Roger L. Martin in their 2013 book, published by Harvard Business Review Press; Throughline is not affiliated with or endorsed by the authors, Procter & Gamble or the publisher. Written by Tom Olajide, Founder. Last reviewed September 24, 2026.