The Pyramid Principle, popularized by McKinsey's Barbara Minto, is a framework for organizing information.
The pyramid has two core principles:
- Synthesize from the bottom up
- Communicate from the top down
Synthesize from the bottom up
The process assumes that you have a clearly defined problem and hypothesis to drive your research.
You can use the principle to organize your data into three levels:
- Lower level: Information, analysis, survey results. In other words, raw data.
- Middle level: Synthesized arguments, themes, or takeaways of a group of arguments. In other words, the key insights or patterns that have emerged from the data. Each insight should be composed only of the elements below it in the structure.
- Top level: The main takeaway from the information you have analyzed—the big insight that sums up everything you have learned.
At each level, strive to be mutually exclusive and comprehensively exhaustive (MECE).
Communicate from the top down
Present your findings in the reverse order in which you developed them. That is, communicate from the top down.
Start with the answer. Remember that people can only hold a few "chunks" of data in their head at any one time.
Consider writing a memo rather than creating a deck, at least to start. writing-is-thinking Begin with the overall recommendation and summarizing the key insights. (This is your topic sentence.)
Then, address each point in turn. Once again, ensure that your ideas are MECE. Each point would be its own paragraph.
Direct storytelling versus indirect storytelling
Usually, starting with the answer (direct storytelling) is the best approach for persuasive writing. Use this approach when dealing with
- friendly clients;
- impatient clients; or
- big-picture, strategic discussions in which there is already some acceptance around the proposed recommendations.
However, if you fear pushback, you may use indirect storytelling to ease the audience into accepting your recommendation. Use the indirect approach for
- controversial recommendations
- hostile audiences
- analytical organizations and personalities
- audiences who love data but are not impatient
