What good strategy is: the marks of good and bad strategy, and ten questions to test yours

The last slide of many strategy presentations reads something like this: grow profitably, delight our customers, be the employer of choice, become digital and sustainable. Nobody in the room objects, and that is the problem. A list nobody could disagree with contains no choice, so it cannot guide a single decision on Monday morning. Good strategy is different, and you can test for it.

Richard Rumelt of UCLA Anderson calls such lists bad strategy. In Good Strategy/Bad Strategy (Crown Publishing, July 2011) and an article in McKinsey Quarterly he describes good strategy as coherent action backed by an argument. This page combines his ideas with those of Michael Porter and of A.G. Lafley and Roger Martin, and ends with ten questions you can put to your own plan.

What is bad strategy? Four marks

Rumelt also warns against template-style planning: filling in the blanks for vision, mission, values and strategies. The result looks like strategy without the hard work of choosing; a long list of priorities is often a polite way of avoiding the choice.

The kernel of good strategy: diagnosis, guiding policy, coherent action

1. 01

   Diagnosis

   An explanation of the nature of the challenge. It simplifies a confusing situation by naming the few aspects that matter most, and says why the problem is hard.

2. 02

   Guiding policy

   The overall approach chosen to deal with the obstacles in the diagnosis. Not a goal, but a direction that rules some options out.

3. 03

   Coherent actions

   Steps that are coordinated with each other and support the guiding policy. Separate projects that are each sensible but do not reinforce each other do not count.

The order matters. Start with actions and you get a project list; start with goals and you get a wish. Start with the diagnosis and you know which actions are needed and which can wait.

Where to play and how to win belong together

A.G. Lafley, then chief executive of Procter & Gamble, and Roger Martin describe strategy in Playing to Win (Harvard Business Review Press, February 2013) as a cascade of five linked choices: what is our winning aspiration, where will we play, how will we win where we have chosen to play, which capabilities must be in place, and which management systems make sure those capabilities exist.

The middle two only work as a pair. A choice of market without a reason to win there is a wish; a way to win without a chosen market is a slogan. Where you play weighs heavily: 80% of the variance in revenue growth is explained by choices about where to compete (source: McKinsey, January 2011).

Martin also separates strategy from planning (Harvard Business Review, January 2014). Costs lend themselves to planning because the company controls them; for revenue, customers are in charge, and no plan makes it appear. A detailed budget is useful, but it is not a strategy.

Be different, and choose what you will not do

Michael Porter made the same point from another angle in What Is Strategy? (Harvard Business Review, November 1996). Operational effectiveness, doing the same things better than rivals, is necessary but not enough, because competitors copy improvements quickly. Strategy means creating a valuable position that differs from your rivals' and rests on a different set of activities.

Two ideas hold such a position together. A trade-off means that more of one thing requires less of another: a firm that promises same-day delivery of everything cannot also run on the lowest stock. Fit means the activities reinforce each other, so a rival cannot copy one piece and get the result. Hence Porter's line that "the essence of strategy is choosing what not to do".

Strategy also emerges along the way

Not every strategy is written in advance. Henry Mintzberg and James Waters described deliberate and emergent strategies as the two ends of one scale (Strategic Management Journal, July 1985): what a firm actually does is partly planned and partly a pattern that grows out of many decisions. Many successful small firms never wrote a strategy, yet follow a clear pattern.

Good strategy therefore fixes the main choices and the assumptions behind them, and leaves room to learn. How much room depends on your market. Where customers and competitors are predictable, analysis and a multi-year plan work well. Where the market moves fast or you cannot shape it, short cycles, small experiments and clear signals for changing course work better.

Ten questions to test your strategy

Chris Bradley, Martin Hirt and Sven Smit of McKinsey proposed ten tests for a strategy; in a survey of 2,135 executives, few strategies passed more than three of them (source: McKinsey, January 2011). The questions below combine their tests with Rumelt, Porter and Lafley and Martin. Answer each with yes or no, and count a yes only if you can point to where the answer is written down or visible in the budget.

• Diagnosis: can you state the main challenge or opportunity, and why it is hard, in two or three sentences?
• Guiding policy: is there an approach that follows from the diagnosis and rules options out?
• Where to play: are customers, markets and products specific enough that staff know which request does not fit?
• How to win: is it clear why those customers choose you over their best alternative?
• Different, not just better: does your advantage come from a different way of working rather than from working harder?
• What you will not do: have you written down which customers, products or activities you deliberately leave alone?
• Coherence: do the actions reinforce each other, or are they separate projects?
• Resources: do you know which capabilities you need, and do money and people visibly move towards the chosen route?
• Assumptions: have you stated what must be true for the strategy to work, and which signal will make you change course?
• Plain language: could someone who was not in the room retell the strategy without buzzwords and without just listing goals?

Fewer than five yes answers puts you in large company. Start with the diagnosis: most of the other answers follow from it.

From weak to strong: three examples

Fictitious Firmfocus examples.

The strong versions start from a problem, make a choice someone could object to and say what stops. Weigh such options with evaluate and choose, and see acquisitions and alliances for what the building materials division might do with the countries it leaves.

Pitfalls when you test your strategy

• Marking your own homework: the people who wrote the strategy find it clear. Ask an outsider, such as your accountant or a lender, to answer the ten questions too.
• Intentions as answers: "we will focus more" is not a yes. Look for the choice in the budget, the staffing plan or the customers you turned down.
• One test, then never again: assumptions age. Repeat the test at every yearly review and whenever a warning signal appears.
• Fixing the wording instead of the choice: a sharper sentence about a vague choice is still fluff.

Good strategy in a small business and a larger organisation

This page gives you the test; the other pages in strategic choices give you the models for the choices it asks for, from Porter's generic strategies to Blue Ocean Strategy. Capture the result in a one-page strategic plan. Every topic is in the strategy overview.