One page or nothing

The page limit is not a formatting preference — it forces the ranking that makes the document useful, because everything that will not fit is by definition not what the decision turns on.

Count what the rubric currently holds. Nineteen questions across four axes, each with a threshold you want to state. A cover block. A findings table. A remediation estimate with an assumptions list underneath it. A row for what you could not see.

Written at the level of detail you would want, that is not one page. It is four or five, and every line on all five is defensible. Which is exactly the problem: a document where every line is defensible and nothing is ranked hands the ranking to the reader, and the reader cannot do it. They do not know which of your twelve true observations is the one that changes the price.

Nobody published this rule

Before you defend the format out loud, be honest about where it comes from. No source behind this course states a page count for a technical diligence report. The search covered the published methodology pages this course cites — the data-room vendors, the diligence firms, the checklist repositories, the field guide written for the CTO on the other side of the table. None of them names a length. The two that publish worked report material, Papermark and Peony, both show something longer than a page. Both sell data-room software, which is to say both sell the container that longer documents go into.

So the one-page limit is this course’s own design argument. It is not received practice, and presenting it as received practice is checkable in about a minute by anyone who has run a deal. Say it the way it is: I gave myself a page because of what the limit does to the document, and then be able to say what that is.

A limit is a ranking device

Here is the mechanism, and it has nothing to do with attention spans.

On an unbounded document, findings do not compete. Each one is added because it is true, and truth is a low bar that everything you noticed clears. On a bounded document, every finding competes with every other finding for the same fixed space, which forces a comparison you would otherwise never make: is this line worth more to the decision than the line it would push off? You cannot answer that without having a decision in mind, which means the limit drags the reader back into the room every time you use it.

The second effect is on you rather than the document. An unbounded report never makes you defend a cut, so you never find out which of your findings you were keeping out of attachment. A bounded one makes every omission a decision with a reason, and the reasons are the design rationale you will be asked for out loud.

Work out your own budget

The page has fixed furniture and a variable part, and confusing the two is how people end up shrinking the type instead of cutting.

The furniture is everything that has to be there whatever you found: the scope statement, the cover block, the four axis summaries, the remediation estimate with its assumptions, and the what I could not see row. None of that is negotiable. The scope statement is what stops four axes reading as a verdict on the whole company. The assumptions block is what makes the estimate correctable. The access row is the first thing an experienced reader looks for, because it tells them how much weight the rest can carry.

Whatever space is left after the furniture is your finding budget. Do not guess it — lay out the furniture first, on the actual page, and see what is left. The number that comes back is usually smaller than the number you would have guessed, and it is the only honest input to the ranking.

Retrieval check

You have eleven findings and room for six. What rule decides which six?

Check your answer

Not severity, and this is the trap, because severity feels like the obvious answer. Rank by how much each finding moves the decision. A finding that changes a price, a condition of closing, or a 100-day item outranks a more alarming one that changes none of the three, however much worse it sounds.

Two tie-breakers that pull in useful directions. A finding whose cost you can state confidently beats a bigger finding whose cost you are guessing at, because the second one will collapse the first time someone with more information asks how you got there. And a not assessed mark on something material beats a low-value green, because telling the reader what you could not reach is worth more than telling them about something nobody has to fund.

The five you cut do not evaporate. They go on a cut list, with a one-line reason each, and the cut list travels with you into the room without going on the page.

The cut list is part of the deliverable

There is an obvious cheat available here, and it is worth naming so you do not take it: appending three pages of backup and calling the first one the summary. That is not a one-page document. It is a four-page document with a cover sheet, and it fails in exactly the way the limit was meant to prevent, because nothing had to compete.

The honest version keeps the boundary sharp. The page is what gets read. The cut list is what gets asked for. You bring it, you can produce it instantly, and every entry on it says in one line why it did not make the page. That is a different artifact with a different job, and the difference matters in the room: a reader who asks “did you look at X” and gets a straight answer about why X was cut trusts the page more, not less.

There is also a diagnostic buried in the cut list. If everything you cut came off the same axis, one of two things is true. Either that axis is not earning its place on the page, or your evidence for it is thin and the findings were weak before you ranked them. Both are worth knowing before somebody else notices.

Do not shrink your way out of the limit

Every formatting escape hatch is available to you and all of them defeat the purpose. Smaller type, tighter leading, a second column, a landscape orientation, an appendix, a footnotes block, a linked document. Each one lets a finding onto the page without anything being pushed off, which means the ranking never happens.

The test is simple and you can apply it to your own draft: did adding this line force something else off? If not, you are not working to a limit. You are working to a layout.

What the page owes the reader who was in no meeting

The vendor sources are useless on format and genuinely useful on audience. Peony — and read this with the bias visible, because Peony sells the software that data rooms run on — describes the standard as lender-grade: a report where material findings are evidenced rather than asserted, because the person reading it was in none of the conversations that produced them.

That is the constraint that makes a one-page limit hard rather than convenient. You are not compressing by leaving out the evidence. Each surviving line still has to carry what it rests on, or it becomes an assertion from a stranger. Compression has to come from having fewer lines, not from thinner ones, and that is the whole reason the ranking is unavoidable.

Hands on

Cut the rubric to a page

Done when: RUBRIC.md contains a one-page draft laid out with its furniture first, and a Cut list section where every cut item carries a one-line reason.

Open learning/ux-diligence/RUBRIC.md and lay out only the furniture on a single page: scope statement, cover block including the “what I could not see” row, four axis summary lines, and the remediation estimate with its assumptions block. Set the type at a size you would actually read across a table, not the smallest size that fits.
Measure what is left. Count the lines the findings can occupy. Write that count down in the file as your finding budget, and note that you measured it rather than chose it.
List every finding you currently have as a candidate, unranked. Then rank them by one question only: how much does this change the price, a condition of closing, or a 100-day item?
Fill the budget from the top of the ranking. Stop when it is full, even if the next item is true and annoying to lose. Especially then.
Add a Cut list section to the file with everything that did not fit, one line of reasoning each. “True but nobody has to fund it” is a reason. “Ran out of room” is not — if that is what you wrote, you have not ranked, you have truncated.
Read the cut list back and check which axis it came from. If it is mostly one axis, write two sentences on which of the two explanations applies: the axis is not carrying its weight, or your evidence there is thin.
Bring the page and the cut list into the chat together. The argument worth having is about the last item you kept and the first one you dropped, because that boundary is where your ranking rule is actually visible.

What this does not cover

A ranked page still says nothing about whether the colours printed on it mean anything. Right now the thresholds in the rubric are written for an audience of one, and a colour only one person can reproduce is a preference with a swatch next to it. The red/amber/green test lesson is about writing a threshold before you score, so that a second reader given the same evidence lands in the same place — and about refusing the numeric precision that a four-colour instrument has not earned.

The estimate that sits under the axis rows gets its own treatment in the remediation-in-weeks lesson, which is about what has to be printed beside a number for the number to be correctable. After that comes the finding management will dispute, and then the lesson that runs the whole instrument against a real product you can only see from the outside.