Good Strategy, Bad Strategy — Book Report

About a month ago I started working on a project where I had to help formulate a technical strategy for a very large project, so I thought this would be a good time to learn how to think about effective strategies. I found a lot of resources (particularly from Will Larson’s wonderful staffeng.com website), but a few of the blog posts recommended the book Good Strategy Bad Strategy: The Difference and Why It Matters so I decided to read it. Overall I enjoyed the book and learned a few things from it. I recently decided that I’m going to try and write book reports just to further help me internalize the key points.

There’s no particular structure to bullet points below, as I ossicilate between summarizing the key takeaways from the book and sharing my own personal commentary.

• The beginning of the book focuses on characteristics that make certain (in my opinion most) strategies ineffective. The author Richard Rumelt characterizes bad strategies having the following qualities: fluff, failure to face the challenge, mistaking goals for strategy, and bad strategic objectives.
• Fluff is pretty straightforward, so I’m not going to spend much talking about this one. But if I were to define it, it’d be “words that make your eyes roll.”
• Failure to face the challenge means when the strategy doesn’t acknowledge the real underlying problem or obstacle that the organization faces. In one of his case studies, Rumelt describes International Harvester’s strategy as “high-sounding sentiments together with plans to spend more and somehow ‘get better.’” Another key quote that best captures this concept (bolded mine): “If you fail to identify and analyze the obstacles, you don’t have a strategy. Instead, you have either a stretch goal, a budget, or a list of things you wish would happen.”
• Mistaking goals for strategy is something I see in a lot of my past companies (e.g. grow sales by X%) yet fortunately I don’t see too often here at Stripe (one of the major reasons I wanted to join Stripe was to see how effective organizations run, and I think I made the right decision here). I suspect that many startups sets high level goals, mistaking it for strategy.
• Rumelt presents two categories that fall under bad strategic objectives: a) dog’s dinner objectives and b) blue-sky objectives. Dog’s dinner is something I’m guilty of committing in the past: simply presenting a todo list. Blue sky objectives simply mean setting impossible/vague tasks. This is common in the startup-world, since you don’t form startups without having some irrationality in you.
• Rumelt presents some potential reasons why people choose bad strategy. One of them is the unwillingness/inability to make a tough decision. This manifests in trying to tackle too many things at once, stretching the organizatin thin and diluting the product quality or focus. I’ve also seen leaders punt making tough decisions because they’re afraid hurting others’ feelings.
• Rumelt talks about the history behind “template-style strategy”, which was really fascinating. I won’t re-hash the history (it’s a long but good read), but template-style strategy is “a reduction of charismatic leadership to a formula: (1) develops or has a vision, (2) inspires people to sacrifice (change) for the good of the organization, and (3) empowers people to accomplish the vision.” It seems like many corporate strategies that are publicly available follow this formula (and in startups, this is pretty much the status quo).
• While having a vision isn’t as valuable as having a solid strategy, I still think writing down a long term vision is useful. It’s helpful to know specifically what the ideal end-state should be. I use it as guiding posts as I’m thinking about what are the specific obstacles that are preventing the higher level goals and whether the problem I’m trying to solve actually aligns with that vision. If there’s no alignment, then perhaps the vision isn’t realistic or the right one. Adjust the vision, rinse and repeat.
• Rumelt provided ~100 year historical context behind the New Thought movement, a major contributor to the current prevalent thinking that strong leadership is having immense self-confidence and having a “winning” attitude at all times (e.g. “We will defeat the enemy!”). I’ve seen many TED talks and self-help books talking about self-actualization and pushing away negative thoughts, and this is especially prevalant in the startup culture, so it’s fascinating to learn about the history behind the current leadership culture.
• Chapter 5: The Kernel of Good Strategy. Finally, the good stuff! Good strategy has at the very minimum three elements: 1) a diagnosis of the challenge/opportunity, 2) a guiding policy, and 3) a set of coherent actions designed to carry out the guiding policy.
• Rumelt makes the point that having the proper diagnosis is a critical (if not the most critical) piece of a good strategy. Identifying the problem enables one to look back in history (whether it’s from personal experience, blogs/books, or interviews with other people) to pattern match similar problems and apply a past solution that worked. In other words, how you frame the problem can quickly constrain the solution space. Effectively diagnosing the root problem is a hard challenge since there are many ways to interpret the underlying problem, and having an incorrect diagnosis can lead down a costly and wrong path.
• The diagnosis element reminds me of the Situation/Complication elements from the SCQA, a tremendously valuable writing framework that I learned from Jeff Weinstein at Stripe. The biggest value I got from SCQA was to consciously think and write down the actual problem I’m trying to solve vs trying to prescribe a solution upfront. Many times when I write down what I think the problem is, I realize either a) I have a huge knowledge gap in the problem space (and I should talk to more people to better understand the problem) or b) I completely mis-diagnosed the actual underlying problem (and I should talk to some more people). If there’s one thing that has helped me in my career, it’s dedicating time to writing out my perceived understanding of the world.
• Guiding policy is a really nuanced but, in my opinion, really insightful concept. Rumelt best describes it as the following (bolded mine): “Good guiding principles are not goals or visions or images of deseriable end states. Rather, they define a method of grappling with the situation and ruling out a vast array of possible actions.” He then goes into examples of comparing visions to guiding policies e.g. Wells Fargo has the corporate vision of “satisfy all our customers’ financial needs [and] help them succeed financially”, but their guiding policy was to cross-sell their financial products as much as possible (although in retrospect this may had the unintended consequence of fostering a toxic culture, resulting in Wells Fargo’s illegal activities up until 2016).
• The final element of a good strategy is defining what actions to take. They should be “consistent and coordinated” with one another, where coordination is not “ad hoc mutual adjustment” but rather “coherence imposed on a system by policy and design.” While coordination can be powerful, it can make the organization overall less efficient, and so the potential gain must be worth the cost.
• Explicitly writing out goals (and non-goals) is somewhat similar to writing out the guiding principles. It helps reduce the solution space. However, the key difference is that goals/non-goals doesn’t necessarily suggest what general actions can be taken (although it can constrain the solution space so much that it’s obvious what the only path forward is).
• In part 2 of his book, Rumelt focuses on “sources of power” for good strategies. There are some gems in this part and some that confirms existing beliefs (which is reassuring!). I won’t focus on all of them, but I’ll talk about the ones that I found insightful or resonating.
• Rumelt talks about setting proximate objectives, which is basically setting an objective that you have high confidence in achieving. When the situation/environment is really complex and it’s hard to predict the future beyond a certain time horizon and can cause analysis-by-paralysis. By setting a proximate objective and committing to it, it unblocks the organization to come up with some near-term goals that directionally goes in the right way. This chapter resonated with me, and I recently went through an excercise at Stripe where proximate objectives was a really useful tool.
• Chain-link systems describe interlinked processes, organizations, or competitive strengths that collectively brings more value than the sum of the individual parts. Interestingly, sometimes investing in just the weakest link by itself can be more detrimental than doing nothing at all. This is something I had never considered deeply before, so I was really excited to learn about this.
• Having intentional design when constructing a strategy is a “configuration of resources and actions that yields an advantage in a challenging situation.” The tradeoff, however, is that there’s an overhead cost with all the coordination. Additionally, if you had better resources (e.g. more engineers, people with specialized expertise, large streams of revenue, etc) then coordinating all your resources and actions may not be necessary. On the flip side, having better resources can make organizations become complacent and masks the need for a coherent strategy.
• In chapter 11 “Dynamics”, Rumelt talks about an “attractor state”, which describes an ideal state where the buyers’ needs and demands are efficiently met with the lowest cost possible. He used IP addresses as an example. In the 1990s, many networking hardware companies created proprietary network protocols, which caused friction for ITs/businesses as they’re trying to integrate different networking solutions. The attractor state is a novel concept for me, and I’m storing this in my toolbox for future use.
• Part 3 is more brief and less substantive than the first 2 parts, but in many ways it’s the most important section: tactics on how to actually form good strategy. The tl;dr? At the end of the day, forming strategy is no different than the scientific process: you’re making a highly educated guess based on data/observations that’s readily available (and your experiment is the execution of that strategy). So to think as objectively and clearly as possible, explicitly writing down your thoughts like the challenges, competitive advantages, goals/non-goals, guiding principles, etc. forces you to take a stance, which can then later be reviewed as you get more data/feedback on the strategy. Rumelt also suggests trying to actively tear down your strategy. One tactic he suggested was to think of different personas in your life (e.g. Steve Jobs, your manager from your last company, etc) and try to think about the problem/solution-space from their perspective.
• In the last chapter, he finally ends with “Keeping Your Head,” which talks a about how to not not let your biases cloud your judgement. It reminded me of Thinking, Fast and Slow, another fantastic book.