The Pyramid Principle by Barbara Minto - Book Summary

41SXHdBJ-+L._SX310_BO1,204,203,200_.jpg

This is my summary of the book ‘The Pyramid Principle: Logic in Writing and Thinking‘ by Barbara Minto. My notes are informal and tailored to my own interests at the time of reading. They mostly contain quotes from the book as well as some of my own thoughts. I enjoyed this book and would recommend you read it yourself (check it out on Amazon).


Principles of effective document structure

  • The order in which you present your thinking makes your writing clear or unclear: 

    • If a person’s writing is unclear, it is most likely because the ordering of the ideas conflicts with the capability of the reader’s mind to process them. 

    • You should impose the proper order: create a comprehensive structure that identifies the major ideas and their flow, and organize the minor ideas to support them. 

    • The key skill is to be able to recognize which are your major and which your minor ideas, and work out their relationships within the structure. 

  • Every document you write will always be structured to support only one single thought - the one that summarizes your final set of groupings. 

    • This should be the major point you want to make, and all the ideas grouped underneath - provided you have built the structure properly - will serve to explain or defend that point in even greater detail. 

    • You can define in advance whether or not you have built the structure properly by checking to see whether your ideas relate to each other in a way that would permit them to form pyramidal groups. 

  • Ideas at any level in the pyramid must always be summaries of the ideas grouped below them. 

    • Double check the origin of the grouping to make sure it’s MECE (i.e. that its order reflects a process, a structure, or a classification). 

    • Look at the kind of statement you are making: 

      • An action statement, telling the reader to do something. Summarize the action ideas by stating the effect of carrying out the actions. 

      • A situation statement, telling the reader about something. Summarize the ideas by stating what’s implied by their similarity to each other (inference). 

    • Be specific. Poor examples of summaries: 

      • The company should have three objectives. 

      • There are two problems in the organization. 

      • We recommend five changes. 

    • There are never more than four points in a chained deductive argument, and never more than five in an inductive one. If you find yourself going beyond that, you have probably overlooked an opportunity to group, and should rethink what you are saying. 

    • You cannot simply group together a set of ideas and assume your reader will understand their significance. Every grouping implies an overall point that reflects the nature of the relationship between the ideas in the grouping. You should first define that relationship for yourself, and then state it for the reader.

  • Ideas in each grouping must always be the same kind of idea. A shortcut in checking your groupings is to be sure that you can clearly label the ideas with a plural noun (e.g. recommendations, reasons, problems, changes to be made, …) 

    • If the first idea in a grouping is a reason for doing something, the other ideas in the same grouping must also be reasons for doing the same thing. 

    • If the first idea is a step in a process, the rest of the ideas in the grouping must also be steps in the same process. 

    • If the first idea is a problem in the company, the others in the grouping must be related problems.

  • Ideas in each grouping must always be logically ordered. There must be a specific reason why the second idea comes second, and cannot come first or third. 

    • The order you choose should reflect the analytical activity that your mind performed to create the grouping, and there are only four possible logical ways in which to order a set of ideas.

    • If it was formed by reasoning deductively, the ideas go in argument order (major premise, minor premise, conclusion): 

      • In deductive groupings, logical order is the order imposed by the structure of the argument. 

    • If it was formed by working out cause-and-effect relationships, the ideas go in time order, chronologically (first, second, third): 

      • Whenever you make statements in writing that tell the reader to do something, you do so because you believe the action will have a particular effect. You have determined in advance the effect you want to achieve, and then identified the action necessary to achieve it. 

      • When several actions are together required to achieve the effect (e.g. 3 steps to solve a problem), they become a process or system. 

      • A grouping of steps that represents a process or system always goes in time order. 

    • If it was formed by commenting on an existing structure, the ideas go organized structurally (Boston, New York, Washington): 

      • For example org charts, or picturing the structure of an industry. 

      • When you divide a whole into its parts, whether it be a physical whole or a conceptual one, you must make sure that the pieces you produce are MECE (mutually exclusive of each other, collectively exhaustive in terms of the whole). 

    • If it was formed by categorizing, it goes in order of importance, comparatively (first most important, second most important, etc): 

      • Whenever you say that a company ‘has three problems’, you are not speaking literal truth. The company has many problems, of which you have classified three as being noteworthy in some way compared to the others. 

      • Make sure to group things together to avoid having more than 3-4 items. If you find 8 problems, group them by category of problem.

  • Any point you make must raise a question in the reader’s mind, which you must answer on the line below (horizontally): 

    • Your document should be a question/answer dialogue.

  • Always give the summarizing idea before you give the individual ideas being summarized. 

    • Bad: 

      • “I was in Zurich last week - you know what a conservative city Zurich is - and we went out to lunch at an outdoor restaurant. Do you know that within 15 minutes I must have seen 15 people with either a beard or a moustache.” 

      • (now I have given you a piece of information and without realizing it you will automatically make some assumptions about the reason for my giving you that information. You might think, ‘She’s talking about how unconservative Zurich is getting’ or ‘She’s going to compare Zurich with other cities’ or ‘She’s hung up on beards and mustaches.’ Regardless of what reaction you have, the point is that your mind is waiting for further information on one of those same subjects, whatever it turns out to be.) 

      • “And you know, if you walk around any New York office you can rarely find even one person who doesn’t have sideburns or a moustache.” 

      • (now what am I getting at? I seem to be comparing not cities as such, but cities in which we have offices; and instead of just beards and moustaches, I seem to be including all manner of facial hair. ‘Probably she disapproves of the new style. Or maybe she’s going to compare the style in various offices. Or maybe she’s surprised at the amount tolerated in the consulting profession.’) 

      • “And of course facial hair has been a part of the London scene for at least 10 years.” 

      • (‘Ah,’ you think, ‘at last I see what she’s getting it. She’s trying to make the point that London is ahead of all other cities.’) 

    • Better: 

      • “You know it’s incredible to me the degree to which facial hair has become such an accepted part of business life. 

        • In Zurich…

        • In New York…

        • And of course in London…” 

  • Use headings to highlight your structure: 

    • If you formulate your headings properly, they will stand in the table of contents as a precis of your report. 

    • Never use only one of any element. Since the headings indicate levels of abstraction in the pyramid, you can never have only one item at each level. You can never have only one major section, or one subsection, or one numbered paragraph, or one dash point. 

    • Show parallel ideas in parallel form. Use the same grammatical form for the wording of each heading. Example: 

      • Appoint a full-time chief executive. 

        • To coordinate activities

        • To effect improvements

      • Establish clear lines of authority

        • Regrouping hotels by support needs

        • Assigning responsibilities for overseas operations

        • Removing boards from the chain of command

    • Limit the wording to the essence of the thought. Headings are meant to remind, not to dominate. Make them as concise as possible. ‘Appoint a full-time chief executive to provide clear central authority’ is too long. 

    • Don’t regard headings as part of text. They are often not read carefully, and you cannot depend on them to carry your message. Make sure your opening sentence indicates that you are turning to a new topic. Your entire document should be able to be read as a smooth-flowing piece without the heading.

Creating your narrative

  • Make sure you know what question it is you are trying to answer. Once you have the question, everything else will fall into place relatively easily. Include only the information that will be relevant to the reader’s question. 

  • It is generally easier to start at the top and work down because you begin by thinking about the things that it is easiest for you to be sure of - your subject and the reader’s knowledge of it, which you will remind them of in the introduction. 

  • Step 1: Write down the subject you are discussing. 

  • Step 2: Decide the question. Visualize your reader. To whom are you writing, and what question do you want to have answered  in their mind about the subject when you have finished writing? 

  • Step 3: Write down the answer. 

  • Step 4: Identify the situation. Take the subject, move up to the situation, and make the first noncontroversial statement about it you can make. What is the first thing you can say about it to the reader that you know they will agree is true - either because they know it,or because it is historically true and easily checkable? 

  • Step 5: Develop the complication. Say to yourself, ‘So what?’ Think of what happened in that situation to raise the question. Something went wrong, perhaps, some problem arose, or some logical discrepancy became apparent. 

  • Step 6: Recheck the question and answer. The statement of the complication should immediately raise the question you have written down. If it does not, then change it to the one it does raise. Or perhaps you have the wrong compilation, or the wrong question, and must think again. 

  • Example: 

    • Subject = Customer requested a change

    • Situation = They requested change

    • Complication = You asked if it makes sense

    • Question = Does it make sense? 

    • Answer = Yes, it is a good idea

    • New question = Why? 

    • Subpoints: will give us all the information we need; will increase our cash flow; will reduce our workload

Use inductive reasoning at higher levels, and deductive reasoning only at paragraph level

  • Deduction: 

    • Deduction presents a line of reasoning that leads to a ‘therefore’ conclusion, and the point above is a summary of that line of reasoning, resting heavily on the final point. 

    • Example: 

      • The purpose of the monopolies law is to stimulate production and distribution 

      • → The union monopoly over manpower stops production and distribution

      • → Therefore the unions should be controlled by the monopolies law 

  • Induction: 

    • Induction defines a group of facts or ideas to be the same kind of thing, and then makes a statement (or inference) about that sameness. The deductive points derive from each other, the inductive points do not. 

    • In inductive reasoning the mind notices that several different things (ideas, events, facts) are similar in some way, brings them together in a group, and comments on the significance of their similarity. 

    • Example: 

      • Poland is about to be invaded by tanks: 

        • French tanks are at the Polish border

        • German tanks are at the Polish border

        • Russian tanks are at the Polish border

  • When to use which: 

    • Deduction is a useful way to think, but a ponderous way to write. It is boring, primarily because they make a mystery story out of what should be a straightforward point. 

    • On the key line level, try to avoid using a deductive argument, and strive instead always to present your message inductively. It is easier on the reader. 

    • As a rule of thumb, it is always better to present the action before the argument. Two situations in which the argument might be more important to the reader than the action: 

      • If he is going to disagree strongly with your conclusion, so you must prepare him to accept it. 

      • If he is incapable of understanding the action without prior explanation, so you must give him the reasoning that underlies it. 

    • At the paragraph level, deductive arguments are lovely, and present an easy-to-follow flow. But inductive reasoning is always easier to absorb at higher levels.

How to write captivating introductions

  • Start with a question that’s relevant for your reader: 

    • The pyramid structure permits you to carry on a question/answer dialogue with your reader. This question/answer dialogue cannot be counted on to engage their interest unless the statement that starts it off is relevant to them. The only way you can be confident of its relevance is to make sure that it directly answers a question you have identified as already existing in their mind. 

    • You write primarily to tell people what they don’t know. But a reader wants to find out what they don’t know only if they need to do so. If they have no need, they will have no question, and vice versa. 

    • Your introduction tells the reader what they already know or could reasonably be expected to know about the subject you are discussing, and thus reminds them of the question they have to which they can expect the document to give them an answer. 

  • Always put historical chronology in the introduction. 

    • Simple historical occurrences do not exist as the result of logical thought, and therefore cannot be included as ideas. 

  • Limit the introduction to what the reader will agree is true: 

    • The introduction is meant to tell the reader only what they already know. They are meant to remind rather than inform. 

    • If they don’t already know, make sure that the point being made can be easily checked by an objective observer and deemed to be a true statement. Then your reader will not question its truth. 

    • If you have any children you know that the best stories in the whole world are ones they already know. Consequently, if you want to tell the reader a really good story, you tell him one he already knows or could reasonably be expected to know if he’s at all well informed. 

    • Tell your reader things with which you know they will agree, prior to your telling them things with which they may disagree. Easy reading of agreeable points is apt to render them more receptive to your ideas than confused plodding through a morass of detail. 

  • The length of the introduction depends on the needs of the reader and the demands of the subject: 

    • Generally two or three paragraphs. Can also be as short as a sentence: “In your  letter of January 15 you asked me whether…’ 

    • It must say just enough to remind the reader of their question. 

    • The situation and the complication can be as long as three or four paragraphs, but never more than that. It can’t take very much to remind someone of what they already know. If you find yourself littering the introduction with exhibits, you can be sure that you are overstating the obvious. 

  • Follow the storytelling pattern: situation - complication - question - answer: 

    • The introduction should begin by establishing for the reader the time and place of a situation. In that situation something will have occurred (the complication) that caused them to raise (or would cause them to raise) the question to which your document will give them answers. 

    • Once you state the answer (your main point at the top of your pyramid), it will raise a new question in the reader’s mind that you will answer on the line below. 

  • Where to start the situation: 

    • Make a statement about the subject that you know the reader will agree with because you are telling them something they already know. 

    • If you find can’t make a statement about the subject, then either you have the wrong subject, or you’re starting in the wrong place to talk about it. 

    • Arrange known material in a narrative form. Inspire your readers to ask the question you wish to address. 

    • Examples: 

      • Energoinvest is considering the possibility of exporting alumina from its Mostar plant to Ziar in Czechoslovakia. (Memorandum)

      • Every major health service is beset by increasing pressure on already scarce resources - and the Irish Health Service is no exception. (Report)

      • If we accept the common usage of words, nothing can be more readily disproved than the old saw, ‘You can’t keep a good man down.’ (Book) 

  • What’s a complication: 

    • It describes an alternative to a stable situation, rather than a problem per se, although sometimes the alteration is a problem. 

    • Examples: 

      • Something went wrong (--> what do we do?) 

      • Something could go wrong (--> how can we prevent it)

      • Something changed (--> what should we do?) 

      • Something could change (--> how should we react?)

      • Here’s what you might expect to find in it (--> do we find it?) 

      • Here’s someone with a different point of view (--> who is right) 

      • In this situation we have three alternatives (--> which one should we take?) 

  • You can choose to arrange the parts of the introduction differently to reflect the tone you want to establish in the document: 

    • Considered (situation - complication - solution): 

      • In recent years, the firm has billed dozens of clients large amounts of money for diversification work. However, as yet no one in the London Office can claim the magnum of champagne available to the first consultant who can demonstrate an acquisition or merger by a client that would not have happened without our efforts. Since our diversification work has increased by 40% in the past 5 years, the time is ripe for a firm development project to determine how we can ensure that diversification studies do bring significant benefits to the clients we serve. 

      • This memorandum outlines the major issues and hypotheses that should be resolved and tested during the project. 

    • Direct (solution - situation - complication: 

      • Our first priority for a firm development project should be one directed toward improving our ability to help clients diversify. In the London Office alone, our work in helping clients find acquisition and merger candidates has increased by 40% over the past 5 years. Yet we cannot point to a single acquisition or merger that would not have happened without our efforts. 

      • This memorandum …

    • Concerned (complication - situation - solution): 

      • To my knowledge, no one in the London Office has yet conducted a single diversification study for a client that has yielded demonstrable results beyond what he could have done for himself. This situation is startling, since our practice in this area over the past 5 years has grown by 40%. We cannot in conscience go on charging clients for work that does not yield significant benefits and maintain our high reputation. I suggest, therefore, that we conduct a firm development project to determine how can make diversification studies an area of our practice that is proven to bring significant benefit to clients. 

      • This memorandum … 

  • Examples of introductions: 

    • Cracking Japanese Markets (New York Times): 

      • “With the strong yen creating an increasingly competitive environment and Washington achieving a conspicuous lack of success in trade negotiations, many executives in the US and Europe have abandoned their search for new opportunities in the Japanese market. Frustration has set in as attempts to become ‘insiders’ have proved futile. Some companies have shifted their attention to emerging markets like China and India, but by ignoring Japan they are making a big mistake.” 

      • S = US/Europen companies frustrated with Japan

      • C = Have shifted attention to China/India

      • Q - Are they making a mistake

    • Visual thinking in the Ice Age (National Geographic): 

      • “For the first 2.5 million years of the archeological record, the only artifacts left by man were strictly utilitarian: stone tools. Then, about 35,000 years ago in Europe a dramatic turning point was reached. In addition to new kinds of stone tools, we find symbolic objects: the first adornments of the body, in the form of beads and pendants, and the first known attempts at painting and sculpture. This cultural explosion occurred at the same time over large parts of western and eastern Europe. Why?” 

      • S = For 2.5 million years we had only stone tools. 

      • C = Then suddenly 35,000 years ago we get art all over Europe. 

      • Q = Why? 

    • Example of a key line: 

      • This memorandum describes the project team approach to identifying and achieving significant profit improvements. It is organized in six sections as follows: …

Some more helpful pointers for better communication

  • Focus on three ideas at most, only one is even better: 

    • The mind cannot hold more than about seven items in its short-term memory at any one time. Some minds can hold as many as nine items, while others can hold only five. (the magical number 7, +/- 2 - George A. Miller)

    • A convenient number is three, but of course the easiest number is one. 

    • When the mind sees the number of items with which it is being presented begin to rise above four or five, it starts to group them into logical categories so that they can be retained. 

  • Avoid headings called ‘Findings’ or ‘Conclusion’. Such headings have no scanning value. 

  • Transitions: 

    • Make transitions unobtrusive yet clear, primarily through picking up the key word or phrase and carrying it forward. 

    • Example: 

      • No single executive has full-time responsibility for directing group affairs. The absence of necessary leadership and coordination for senior operating and staff executives results in … (list of problems) 

      • The problems stemming from a lack of full-time leadership are compounded by overlapping or unwieldy responsibility assignments … 

Problem solving

  • “Solving a problem simply means representing it so as to make the solution transparent.” - Herb Simon 

  • Problem solving begins with problem setting, i.e. identifying precisely what you mean by ‘the problem’ and how you will know when you have achieved a solution. When you identify a problem, essentially what you do is recognize that a particular situation yields a specific result. 

  • The problem is either that you do not like the result (‘Sales are down 10%) or that you cannot explain the result (‘What determines what we think? Is it the structure of the brain? Or genetic makeup?’) 

  • Sequential analysis: When faced with a problem, answer the following questions in order: 

    • What is the problem? 

      • Example: “We’re losing 2 hours of productivity a day. What I want to know is what is causing those lost hours? Or more specifically, how do I eliminate the causes of lost hours?” 

    • Where does it lie? 

      • Keep digging until (1) you are sure you have identified all the parts in the system, (2) you can arrange them in sequential order, and (3) you can clearly show inputs and outputs. 

      • In the example, five possible things could be going wrong: no materials at the machines, broken machines, no operator, poor operator, or poor supervisor. 

    • Why does it exist? 

      • Not all possibilities are likely to prove equally important in solving the problem. You will have to make a judgment call early on about which areas deserve the greatest concentration of effort. 

      • Such judgments can only be based on experience in the industry or in solving similar problems. 

    • What could we do about it? 

      • Thoroughly assess different courses of action. 

    • What should we do about it? 

      • Before you can legitimately advise someone on how to change an undesirable result, you must have defined clearly five things: 

      • (1) The gap between where he is and where he wants to be

      • (2) The structure of the situation that gave rise to the gap

      • (3) The structure of its underlying processes

      • (4) The alternative ways the structure could be changed

      • (5) The changes required to accommodate the alternative you choose


Click here to browse more book summaries.