Playing to Win, summarised by practitioners.
The 2013 book by A.G. Lafley and Roger Martin, built on a decade transforming P&G, makes one central claim: strategy is a small set of integrated choices, not a plan, a vision or a process.
Most of what companies call strategy is one of the three things Martin says it is not. A plan is a list of actions with dates, and a list commits you to effort rather than to a position. A vision is a destination with the journey left out. A process is a calendar. Playing to Win replaces all three with something less comfortable: five questions, answered as one coherent portrait, where every answer forecloses alternatives. Choice, with consequences.
The engine: five choices, one portrait
The framework's engine is the Strategy Choice Cascade: what winning looks like, where to play, how to win, the capabilities that make the way of winning real, and the management systems that keep those capabilities alive. Five boxes, but the number that matters is one: the test of a cascade is integration, whether the five answers form a single argument in which each depends on the others. Five well-answered boxes that do not need each other is five decisions, not a strategy.
The defining demand
The book's title is its sharpest edge. Martin requires that leaders choose to win, not merely to participate, because in most markets the returns to being distinctively better compound while the undifferentiated slide slowly into commodity competition, the winner-takes-most dynamics that make aiming for "solid" the riskiest position on the board. The opposite posture has a name in our library, Playing to Play, and it is the most common strategy in the world: serve the market, improve steadily, avoid hard choices, hope.
The proof case
The book's evidence is the Olay case study: a dying brand rebuilt into the world's number one skincare business, not by heroics but by disciplined cascade choices, a narrowed field, a specific theory of advantage, and capabilities built for exactly that theory.
How this connects to OKRs
The seam with our OKR methodology is diagnostic in both directions. If a company's OKR portfolio is entirely operational, or its strategy is implicit, you are reading Playing to Play in goal form: bets with no theory of advantage behind them. The fastest test to run on any strategic statement is the Opposite Test, and the distillation of the whole worldview is Martin's Three Laws.
Our opinion: cascades die in box five, not box one
Here is where we part company with how the book is usually taught. Almost every cascade we are handed is well-designed. The winning aspiration is sharp, the where-to-play is narrowed, the how-to-win has a real theory. Then it dies, and it dies in the fifth box, enabling management systems, the one leaders treat as an afterthought at the offsite. Boxes one to four are a portrait. Box five is the only box that runs on a Tuesday. Our claim, and it is falsifiable: a cascade's survival is set almost entirely by box five, and the strategy you cannot see in this quarter's OKRs is a strategy you did not install. OKRs are box five, done in public, on a cadence, with grading that tells you whether the theory of advantage is proving true.
The receipt
A fast-scaling insuretech, fresh from a significant Series C, had the choices but not the home for them. We refocused its OKRs from three company objectives and nine key results to one objective and three in a single quarter, and fixed the metric foundation underneath so the cascade's how-to-win became something a team could read on a dashboard. The choices did not change. The place they lived did. Read it in full in the insuretech case study.
The strategy library here is free and always will be; what we license is the machinery that carries these choices into quarterly execution and keeps them honest.
One line to keep: A cascade dies in box five or not at all.
Our synthesis of published thinking, Roger Martin and A.G. Lafley, Playing to Win (2013), and Martin’s subsequent writing, sources credited. Read the originals: they’re excellent.
Is Playing to Win compatible with OKRs? +
They are the same argument at two altitudes. The cascade decides the few integrated choices; OKRs are the fifth box, the enabling management system that carries those choices into a quarter and grades whether the theory of advantage is proving true. A cascade with no OKR home is a portrait; OKRs with no cascade behind them are busywork.
Why do most strategy cascades fail? +
Not at design, at execution. The first four boxes get the offsite attention; the fifth, enabling management systems, is treated as an afterthought, so the choices never reach a Tuesday. The tell is simple: if you cannot see the strategy in this quarter's OKRs, it was never installed.
Reading about method is not the same as running it. We install this system and build the capability that stays.
Execute your Playing to Win choices →