Read Dense Work Documents With an Extraction Sheet, Not a Highlighter

You open a 40-page proposal because you need to comment on it in tomorrow’s meeting. An hour later, several passages are highlighted, five tabs are open, and you can remember that something on page 17 mattered. But you cannot yet say what the proposal asks for, what could go wrong, or what decision is needed.

The problem is not necessarily that you read too slowly. Highlighting preserves the location of information, but it does not turn information into a usable output. For dense work documents, make a small extraction sheet before you start reading. It gives every note a job: answer a question, identify a decision, flag an assumption, or point to a specific place worth revisiting.

This method is especially useful when you need to digest material that was written by someone else for a practical purpose: a project brief, vendor proposal, internal policy, research summary, handoff document, or planning deck exported as a PDF. It is not meant to replace close reading when you must verify every detail. It is a way to avoid finishing a document with a pile of marks and no clear next move.

Start with the output, not the document

Before page one, write one sentence describing what you must be able to do after reading. Make it concrete.

• “I need to tell the team whether this plan has enough detail to begin.”
• “I need to ask three useful questions in the review meeting.”
• “I need to compare this proposal with our current process.”
• “I need to brief my manager on the decision required and the main tradeoff.”

This sentence narrows your attention without forcing you to predict the conclusion. It also keeps you from collecting interesting facts that do not help with the assignment.

A weak purpose is “understand the report.” Understanding is valuable, but it is too broad to guide your notes. A useful purpose names an action, audience, or decision.

Build a five-part extraction sheet

Use a blank note, a sheet of paper, or a simple document. Create the headings below before you read. Keep the sheet deliberately small: one screen or one page is ideal. A constrained space forces you to summarize rather than transcribe.

Do not fill every section evenly. A clear proposal might need only a line under questions and several under dependencies. A vague proposal might generate mostly questions. The sheet records what the document actually gives you, not what a perfect document would contain.

Skim the structure before you read the details

Spend a few minutes looking at the title, author or owner, date, headings, charts, executive summary, conclusion, and appendices. This is not a shortcut around reading. It creates a map so you can tell whether a detail is evidence, background, a requirement, or a recommendation.

During this pass, fill in only the first heading: purpose and request. If you cannot state the request after skimming, write that down as your first question. A document may be informative yet still fail to make its desired outcome clear.

Also notice what kind of document you are holding. A policy may describe rules and exceptions. A project plan may hide its biggest risks in a timeline or ownership chart. A recommendation may present a preferred option while treating alternatives lightly. Knowing the document’s role helps you read its claims in proportion.

Read in passes that produce different kinds of notes

First pass: capture the spine

Read headings, opening and closing paragraphs, and the first sentence of dense sections. Add short phrases to the purpose, claims, and next-step sections. Do not pause to polish wording.

At the end of this pass, you should be able to explain the document’s basic shape in two or three sentences: the situation, the proposed response, and the action requested. If you cannot, the document may need closer reading—or its central message may be unclear.

Second pass: test the load-bearing claims

Now read the sections that support the recommendation. For each major claim, ask: “What has to be true for this to work?” Put the answer in key claims or constraints.

For example, a plan to automate weekly reporting may depend on source data being complete, consistently named, and available on a predictable schedule. Those conditions matter more than a sentence such as “automation will save time.” They tell you what must be checked before the plan can deliver its promised benefit.

Use a page anchor when a detail may need to be found again. “p. 8: ownership table” is more useful than pasting half the table into your notes. Page anchors keep the extraction sheet compact while preserving a route back to the source.

Third pass: turn friction into questions

When something feels off, resist writing “unclear” and moving on. Name the missing link.

• Instead of “timeline unclear,” write “Who approves the pilot before the stated launch date?”
• Instead of “costs?” write “Does the estimate include the time needed to migrate old records?”
• Instead of “need more info,” write “What happens when the named data owner is unavailable?”

A good question can be answered, assigned, or used to revise a plan. A general concern usually becomes background noise in a meeting.

The most useful reading note is not proof that you covered every page. It is a reliable bridge from a document to the conversation or decision that follows.

Keep observations separate from your interpretation

Dense documents often mix facts, forecasts, preferences, and assumptions. Your extraction sheet becomes more trustworthy when you distinguish what the document says from what you infer.

A simple way to do this is to begin each interpretation with a label such as Implication:, Assumption:, or Question:. For instance:

• Document says: Training is planned for each team.
• Assumption: Team leads will have time to coordinate attendance.
• Question: What is the fallback for people who cannot attend the scheduled session?

This small distinction prevents an easy meeting mistake: presenting your reasonable inference as though it were a commitment the document made.

Use a stopping rule so research does not expand forever

Documents can create new questions faster than they answer old ones. Without a stopping rule, you may chase every unfamiliar term and arrive late to the actual review.

Stop when you can complete these four statements:

1. “This document wants us to…”
2. “Its recommendation depends mainly on…”
3. “Before proceeding, we need an answer about…”
4. “My next action is…”

If the task is a decision with serious consequences, your next action may be a deeper review rather than approval. The extraction sheet still helps because it tells you exactly where deeper review should begin.

It is also reasonable to stop before reading every appendix. Read an appendix when it supports a claim you need to judge, defines a term that changes the proposal, or contains a condition that affects your work. Otherwise, note that it exists and return only if needed.

A realistic example: preparing for a project review

Imagine you receive a plan to replace a shared request inbox with a ticketing workflow. Your purpose statement is: “I need to decide whether my team can adopt this workflow next quarter and identify what we would need to prepare.”

After skimming, you write that the request is approval to begin a pilot. In key claims, you note that the new workflow is intended to make requests easier to assign and track. In constraints, you record that request categories must be agreed upon, staff need access, and someone must own the intake rules.

On closer reading, a chart states that the pilot will include only two request types. Your question becomes: “Which of our requests fall outside those two types, and where will they go during the pilot?” That question is much more productive than saying the plan feels incomplete.

Your final next step might be: “List our five most common request types before the review; bring the routing question to the project owner.” You have not merely read the proposal. You have turned it into preparation for a specific decision.

Common mistakes that make the sheet less useful

Copying whole passages

Verbatim notes feel safe because they preserve nuance, but they hide the point of the passage. Summarize first, then add a page anchor. Copy an exact phrase only when its wording itself is important, such as a requirement or definition.

Making the sheet a second document

If your notes become nearly as long as the report, you have recreated the reading burden for your future self. Limit each claim or question to one or two lines. If it needs more space, it may be a separate analysis task rather than a reading note.

Listing only flaws

Questions are valuable, but a review that captures only problems is difficult to act on. Record what the document does establish: the proposed scope, the intended owner, the stated sequence, or the decision requested. This makes your concerns more precise and fair.

Forgetting the final action

A note without a next step becomes another artifact to reread later. End with a verb: approve, ask, compare, delegate, schedule, verify, or decline. If someone else owns the action, record that too.

Make the method easy to repeat

Save a plain template with the five headings in your notes app or document system. Use it for just one upcoming report or proposal rather than trying to transform every kind of reading at once. Afterward, notice which heading gave you the most value. If you consistently work with policies, you may add “exceptions.” If you read plans, you may split dependencies into “people” and “systems.”

Your next action is small: before opening the next dense work document, write one sentence beginning, “When I finish reading, I need to be able to…” Then create the five headings underneath it. That minute of setup can keep the next hour of reading attached to a real outcome.