Using 6 Page and 2 Page Documents To Make Organizational Decisions

1 Summary

A challenge of organizations is the aggregation of local information to a point where a globally optimal decision can be made in a way all stakeholders have seen their feedback heard and so can “disagree and commit” on the result. This document describes the “6 pager” and “2 pager” document and review meeting process, as a mechanism to address this challenge, as practiced by the document’s author in his time in the EC2 team at Amazon, and then at Two Sigma.

2 Where To Use It

I have seen success with a 6 page document and 60 minute meeting on the following decisions:

  1. Team’s annual plans of what they intend to build vs not build

The major variant I have also seen is 2 pages with 30 minute review; when the decision is smaller in terms of stakeholders, options or impact. That being said, there is nothing magical about 2 pages, i.e., a 3 page document is fine, it just should be expected to take more than 30 minutes to review.

3 Comparison With Other Options

There are a few reasons why for these types of decisions, a document review is preferred to Powerpoint:

Clarity: Written documentation forces the writer to clarify their thoughts before the discussion, and also helps them state them in a way that avoids misinterpretation.

Introvert Inclusion: It levels the playing field for introverts who find it difficult to speak during meeting; often because they are thinking too deeply about issues raised to participate in a fast verbal discussion.

Reader owns rate of information consumption: Written documents can be read at any pace and in any order. They can be pondered. Powerpoint forces a rate of information consumption of the presenter.

Full argument read before conversation starts: Questions that are asked after reading an entire document are informed by the complete picture rather than where the presenter is in the slides. I.e., it avoids the “I address that question in a later slide” problem.

Durability: Written documents are durable. They are records of the decision making process.

Conversation rather than dictation: It is very easy for powerpoint to become the presenter talking the great majority of the time, rather than it being a real conversation amongst all stakeholders.

That being said Powerpoint is fine if it is more information being disseminated than information being pondered, discussed and agreed upon.

Another option for closing decisions is written mechanisms for feedback, e.g., RFCs or “Commenting on Google docs”. These work well when there is agreement on the broad decision, and it is details being worked through. In fact such a mechanism has a nice upside that the documents can be longer, and feedback is asynchronous. But as there is more disagreement, higher personal stakes in outcomes, and in particular a need for “disagree and commitment” from all stakeholders to enable future progress, then the meeting becomes needed to show (or show absence of) social proof of consensus.

All this being said, this is a moderate cost process to write the document, and have the review. So, if a decision can be resolved by a few people at a whiteboard, that is much preferred to a document (although with durability in mind, it is great to capture the outcome somehow, e.g., with a picture from the whiteboard posted in wiki). It is only as stakes get higher, stakeholder groups get larger, or the decision gets more contentious, that a document’s benefits outweigh its costs.

4 Philosophy and Tone

The most important aspect of document is the philosophy behind getting disparate stakeholders to “disagree and commit” consensus. The way it does this is to narratively walk through a background of the area, specifics of the problem, proposed solutions, and their tradeoffs to different stakeholders. When done well, this allows everyone to get in a room, discuss the global picture, and commit on the right decision, even if they still disagree with it.

To achieve this, a neutral “this is the state of the world” tone is crucial:

Own misses: The document should not hide nor be defensive of misses of the authors. If there was a miss that changes how the new decision would be made, dispassionately state it.

Don’t assume other’s misses: Similarly, unless they agree it was a miss and on the lessons being relevant, the document should not be aggressive in highlighting misses of others as mis-decisions.

Disagreements need to be in the document: When there are unresolved disagreements of fact or ideal strategy between stakeholders, they need to be discussed before the document is reviewed and explicitly discussed as options, along with their tradeoffs. This usually means meeting with such stakeholders before the review, to ensure their opinions are understood and can be represented fairly.

5 Document Format and Structure

The format comes from wanting to read and close on a discussion in one hour. This means:

  1. Body is no more than 6 pages.

Following from the philosophy of spending the right amount of time on background material, documents generally have the same coarse structure: 2 pages general background, 2 pages specific background, 2 pages action and plan. For engineering/product/business decision document, an example detailed structure is:

  1. One paragraph, purpose of document. Ideally concludes with recommended solution.

<- End 6 pages->

  1. Appendices: detailed tables, detailed graphs, supporting data.

6 Getting Body Down to 6 Pages

The challenge with this format is packing the needed information into 6 pages. Here is a guide:

  1. Reduce wordiness.

When you have done all the above, but find the body is still more than 6 pages it means that there is too much information needed for a 60 minute review to come to a consensus decision. Ideally that means you tease out separate decisions and author multiple documents with their own rounds of review. This is often the case for a large decision when an author couples requirements, design, implementation and roadmap all into one document, where really the tradeoffs are so large it needs to me made in multiple documents. If you can’t (or don’t want to) do that then the main point is you should not expect a consensus decision to come from a 60 minute review of the document. That is sometimes fine; in some cases stakeholders are fine deferring agreement to the work the author has put in writing the document. In fact there is even a saying for this: the important work is the writing of the document, not the reading of it. The key thing is that type of document, and that type of judgment, differs from what is being laid out here.

7 Review Meeting

Ideally no more than 10 people should be invited to participate in the discussion, otherwise you won’t close the conversation in 60 minutes. On the bottoms-up side this should be the document author and other stakeholders to talk to the detail or strongly affected by the decision outcome. On the tops-down side it should be whoever can give final authority to the document decision; often the intersection of the reporting-line pyramid, but not necessarily, if next levels down are able to “disagree and commit”.

The meeting format is:

  1. Presenter brings hard copies, everyone takes a copy as they enter.

The reason for reading in the meeting is to ensure that everyone has time to read the document. It also ensures they have actually read the document, as opposed to glance over it. That’s also why a paper copy is important, so people aren’t distracted. However, I have found a good portion of people prefer reading the document before hand. Thus best practice is to email out the document 1 business day before the meeting, highlighting that time will be made for reading, and so those that read before should arrive 15 minutes in. Deliberately walking the document in the review is also important; there is a tendency to want to jump to immediately discussing the decision itself. You need to get there, but you want to make sure everyone in the room has had time to ask questions to understand what leads to it.

To close the meeting, the document author should be clear what he sees as next steps. There are generally three:

  1. The document proposal is “disagree and commit” accepted by all stakeholders.

Appendix 1: List of Subjective or Weasel Terms In Documents

a number of
it is said
less than
more than
up to

SVP Core Engineering@ Datadog. Formerly SVP @ Two Sigma, and creator of Nitro @ AWS EC2.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store