I've recommended Good Strategy, Bad Strategy to a few people and it has been mentioned as a useful guide to strategy creation in previous posts. This is probably the best summary I've found of it so far.

I've include the key takeaways and the first part of the more detailed summary below.

Their site also has a few other relevant summaries including Doing Good Better, Algorithms to Live By, So Good They Can’t Ignore You and a couple of 80,000 Hours podcast summaries.

Key Takeaways from Good Strategy, Bad Strategy

• The word “strategy” has been so misused that many people don’t even know what it means.
• Bad strategy fails to face up to the challenge and avoids making hard choices. Bad strategy also fails to bridge the implementation gap — people often confuse strategy with goal-setting and vision.
• The ‘kernel’ of a good strategy has 3 components:
  • Diagnosis. Since a strategy must confront the challenge, the first step is figuring out what the challenge even is.
  • Guiding policy. This is the overall approach to deal with the diagnosed problem.
  • Coherent actions. The kernel doesn’t have to contain every action step, but it has to contain some action to help bridge the implementation gap and those actions should be coordinated to work together.
• Strategy is ultimately about harnessing power and applying it. Sources of power include:
  • Advantages. No one has an advantage at everything, as advantage is rooted in asymmetries.
  • Understanding systems. One reason why strategy is hard is because the connections between actions and results are not always clear.
• Things to watch out for:
  • Attempts to engineer growth. Growth is the outcome of good strategy, not an end to be pursued in itself.
  • Inertia. As organisations grow bigger and more complex, they tend to become resistant to change.
  • Entropy. Over time, organisations naturally become less organised and focused unless they are properly managed.
• Think like a strategist:
  • Strategy is hard and requires analysing unstructured information.
  • New strategies are like hypotheses — success is never guaranteed, but you can test and learn from them.
  • Be aware of your own cognitive limits and biases. The kernel can help with this as it forces you to think carefully about strategy.

Detailed Summary of Good Strategy, Bad Strategy

What is strategy?

For many people, “strategy” has become a meaningless buzzword. But when a word can mean anything, it stops being useful. The word “strategy” comes from military affairs. Rumelt defines it to mean “a way through a difficulty, an approach to overcoming an obstacle, a response to a challenge”.

While business competition is quite different from military battles, there are still valuable business lessons to be learned from military history. Throughout the book, Rumelt uses a variety of business case studies and military examples to illustrate his points.

What is bad strategy?

Bad strategy is not the same as no strategy or a failed miscalculation. Rather, it is the active avoidance of the hard work of crafting a good strategy.

Bad strategy is a way of thinking and writing about strategy that:

• fails to identify or face up to the challenge; and
• fails to bridge the implementation gap.

It often uses fluff — buzzwords, restatement of the obvious — to hide its shortcomings. People with true insight can make a complex subject understandable. People who lack such insight make things complex to mask the absence of substance. Good strategy almost always looks simple and obvious.

Bad strategy fails to face the challenge

Bad strategy glosses over challenges because of concerns that negative thoughts will “get in the way”. But a strategy that fails to identify obstacles isn’t a strategy at all.

Strategy involves focus and focus requires choice. To try and placate everyone, bad strategy spreads rather than concentrates resources. The result is a long list of desirable outcomes or “things to do”. This is particularly common when decisions are made by consensus—universal buy-in indicates that no real choices have been made.

This failure to define the challenge or to make hard choices and trade-offs is bad strategy. It’s a luxury that rich and powerful organisations can get away with, and it’s common in complex organisations with many stakeholders who have conflicting goals.

Rather than focus on a few important items, the group sweeps the whole day’s collection into the “strategic plan.” Then, in recognition that it is a dog’s dinner, the label “long-term” is added so that none of them need be done today.

— Richard Rumelt in Good Strategy, Bad Strategy

Good strategy is as much about what an organisation doesn’t do as it is about what it does. Good strategy focuses energy and resources. Usually, this redirection of resources will make some people worse off, so good strategy often faces resistance.

Example: Intel moves from DRAM to microprocessors 

By 1984 it was clear that Intel could not match its Japanese rivals in producing DRAM (dynamic random access memory). Intel CEO Andy Grove made the hard choice of exiting the memory business and refocused the company on making microprocessors instead.

This was not an easy change. The memory business had driven a lot of research, production, careers, and pride at Intel, and it took over a year to make the change. But it was the right choice. By 1992, Intel became the world’s largest semiconductor company.

Bad strategy fails to bridge the implementation gap

Strategy is not the same as vision, planning or goal-setting. Some examples of bad strategy include:

• To be the graphics arts services firm of choice;
• Grow revenue by at least 20% each year;
• Foster an honest and open work environment; or
• To be a learning community that seeks to serve society by educating the leaders of tomorrow and extending the frontiers of knowledge. 

These are all statements of mere goals or visions. None of them is an actual strategy that specifies how the organisation will achieve its goals.

Strategy should transform vague goals (the organisation’s overall values and desires) into a coherent set of actionable objectives (specific operational targets). A proximate objective — i.e. on that is close enough to be feasible — can galvanise people and help focus efforts. The more uncertain and dynamic a situation, the more proximate your objective should be because it’s harder to see ahead.

Example: Placing a man on the moon 

In 1961, President Kennedy gave a speech calling for the US to put a man on the moon and return him safely to earth by the end of the decade.

This objective was carefully chosen. The Soviet Union had larger launch vehicles at the time, so could be expected to beat the US in any near-term space achievements. But landing a man on the moon would require much larger rockets than either country had. This gave the US an advantage, because it had a larger underlying base of resources.

While the objective may have appeared audacious to laypeople, it was actually quite feasible because US engineers had already developed much of the technology to design and build spacecraft as part of the ballistic missile program.

Every organisation faces situations with overwhelming amounts of complexity and ambiguity. Leaders have the duty to absorb much of that complexity and ambiguity and pass on a simpler, solvable problem. Lower-level units can then look to the high-level proximate objective to work out what their subgoals should be and, in this way, the problem-solving can cascade down the organisation. Many leaders fail badly at this duty.

Inspirational leadership is not strategy

In the mid-1980s, people began to believe that CEOs and other leaders had to be charismatic. The idea was that leaders just inspired and motivated others, and then empowered them to carry out their vision. This idea gained traction because it satisfied leaders’ sense that organisations should “somehow be forced to change” but relieved them from the awkwardness of telling other people what to do.

Charismatic and inspirational vision can indeed motivate people, but inspiration is not a substitute for strategy. Business competition is not just a battle of motivation and grit; it’s also a battle of insights and competencies. While a leader may reasonably call for a ‘push’ to get stuff done, to do their job, they must come up with a strategy worthy of that push.

Planning is not strategy

Many wrongly use the word “strategy” to describe any decision made by the leaders of an organisation, even though many of those are actually planning decisions. Opportunities, challenges, and changes don’t come along in nice annual packages. True strategy work is episodic.

Approving annual budgets or projections for revenues, costs, profits, etc and determining where to allocate resources just maintains the status quo. There’s nothing wrong with planning—it’s just not strategy.

The ‘kernel’ of good strategy

Rumelt sets out the three elements that make up the “kernel” or barebones centre of a strategy:

1. Diagnosis to define the challenge — why you are doing this.
2. Guiding policy for dealing with the challenge — what you are doing.
3. A set of coherent actions to carry out the guiding policy — how you will achieve it.

For example, a doctor’s diagnosis names a disease or pathology. They then choose a particular therapeutic approach as their guiding policy. Lastly, they prescribe medications or lifestyle changes as a set of coherent actions to implement that guiding policy.

1. Diagnosis

Since a good strategy is about finding a way to overcome a challenge, the first step involves figuring out what is even going on. Only after you have a diagnosis can you begin problem-solving.

A good diagnosis simplifies the often overwhelming complexity of reality by identifying certain aspects of the situation as critical. 

— Richard Rumelt in Good Strategy, Bad Strategy

For example, in 2008, Starbucks experienced flat or declining growth and profits. Were the markets saturated? Were consumer tastes changing? Or had competition intensified? Each of these possible diagnoses is a judgment about the critical issues. None of them can be proven to be correct — each is just an educated guess.

A good strategy that insightfully reframes a situation can itself be a source of power. A good strategic diagnosis should also define a domain of action. Sometimes a diagnosis may classify the situation as being a certain type, which lets you draw on knowledge about similar past situations.

Example: Operation Desert Storm 

Towards the end of the first Gulf War, Iraq invaded Kuwait. The US and its coalition members rapidly built up their ground troops but Iraq had a big lead.

US General Norman Schwarzkopf developed a strategy for the ground war involving a two-pronged attack:

• first, air attacks reduced the Iraqi forces by 50%.
• second, while the media focused on the troops in the south, they secretly shifted 250,000 soldiers to the west and north of Kuwait, forming a secret ‘left hook’. When the ground war began, this force slammed into the Iraqis’ flank.

This combined-arms left-hook strategy was very successful. The ground war only lasted 100 hours and the US coalition suffered minor casualties.

The left-hook strategy was hardly groundbreaking. In fact, the US Army publishes field manuals describing its methods and FM100-5 Operations, which was publicly available at the time, even includes a diagram of the left-hook strategy on p 101. Schwarzkopf had simply used one of the US Army’s most basic offensive manoeuvres.

The real surprise wasn’t the strategy itself—it was the fact that a pure and focused strategy was implemented at all.

2. Guiding policy

The guiding policy is the overall approach chosen to deal with the obstacles identified in the diagnosis. Think of the guiding policy like guardrails — it directs some actions and rules out other possible actions, without fully defining the exact steps to be done.

Example: Stephanie’s grocery store 

Rumelt’s friend, Stephanie, owned a corner grocery store and wanted to increase profits. Even in a corner grocery store, there are hundreds of possible changes to make.

The first step was to simplify this complexity with a good diagnosis: Stephanie’s challenge was competition with the local supermarket, which had lower prices and was open 24/7.

Stephanie then divided her customers into several categories (e.g. price-sensitive students or time-sensitive professionals). Framing the problem this way drastically reduced the complexity and helped her see a coherent guiding policy: target the busy professional who doesn’t have time to cook.

That guiding policy suggested a set of actions: Stephanie set up a second checkout stand to handle the burst of traffic at 5pm and repurposed some shelf space to sell high-quality prepared meals to professionals.

Stephanie could not be certain that her policy was the best one. But the guiding policy focused and coordinated her efforts. Without it, her actions would have probably been inconsistent and incoherent.

3. Coherent actions

The kernel of a strategy must contain action in order to bridge the implementation gap. It doesn’t need to contain every action step, because these will change as events unfold. But it has to do more than set the big-picture direction.

When executives complain about execution problems, it’s usually because they’ve confused their job of setting strategy with setting goals. Bringing strategy down to the implementation level can help flush out the conflicts that will result when action has to be taken.

The actions in a strategy should also be coherent — i.e. they should work together to overcome the identified challenge. Coordination alone can be one of the most basic and enduring sources of advantage because a well-coordinated strategy is hard to copy.

Example: IKEA’s highly coordinated system

IKEA’s giant retail stores in suburbs allow for large selections and plenty of customer parking. The catalogues in the stores substitute for a sales force. And not only do the flat-pack designs reduce shipping and storage costs, they also allow customers to take their purchases home right away.

IKEA’s strategy is hardly secret. But more than 50 years on, no one has really replicated it. This is because IKEA has many unusual policies that fit together in a coherent system. Adopting only one of these policies won’t work.

Strategy is an exercise in centralised power

The essence of strategy is coordinated action imposed on a system. It’s an exercise in centralised power to overcome the natural workings of a system. Decentralised decisions work pretty well in many cases but can fail when externalities, principal-agent problems and/or short-term thinking is involved.

For the rest of the summary and a review, here is the link.