September 30, 2026
Takeoff happens a drawing at a time. Pricing does not. You measure eleven sheets and produce eleven material lists, and what has to go out is one set of quantities for one job.
The gap between those two is usually closed by hand, in a spreadsheet, on the afternoon the price is due. It is the least interesting part of the work and it is where a surprising share of the errors get in: a sheet that never made it into the paste, a revision counted twice because both versions were open, a column that lined up on ten sheets and shifted on the eleventh.
This article is about combining material lists from multiple drawings into one consolidated list — what decides whether the result is right, and how to read it.
The same material has to merge. Fifty metres of 110 mm pipe on the ground floor and forty on the first are ninety metres of one item, not two lines a reader has to add up.
Different materials must not merge. A single socket and a twin socket are not the same item however similar the rows look.
You have to be able to see where a number came from. A total of 312 m² is only useful if you can find out which drawings produced it when somebody questions it.
In Planmetry the material summary combines the lists of the drawings you select within a project. Rows merge on the material's name together with its unit — the name being what the material list calls that group of items.
That sentence is the most important one in this article, and it cuts both ways.
Two names for one material give you two rows. "Socket 2G" on one sheet and "2G socket" on another are different materials as far as the roll-up is concerned. You get two rows that each look right and neither of which is the answer. This failure is at least visible: you can see both rows sitting in the list.
One name for two materials gives you one row, silently. This is the dangerous direction. Name and unit are the entire identity — diameter, height, specification, grade and finish are not part of it and do not prevent a merge. If a 15 mm and a 22 mm run are both called "copper pipe", the roll-up adds them together and produces a single figure that is arithmetically correct and commercially useless. Nothing flags the collision as one. The signals are indirect — a property column that comes back blank, or a source count higher than you expected — and you only see them if you look.
So the roll-up is only as good as the naming underneath it, and the fix is in the drawings rather than in the summary — the summary cannot split a row or repair a name. Two habits cover most of it:
Fixing naming after the fact means going back into the sheets and running the summary again, which is exactly the work the roll-up was supposed to save.
One run uses one unit system, so metric and imperial work is not mixed in a single roll-up.
Numeric columns are summed — lengths, areas, counts.
Worth saying once: arithmetic compatibility is not commercial compatibility. Two quantities can be addable and still need to stay apart, because they belong to different phases, floors, options or procurement packages. The summary adds what shares a name; whether they should share a line is your decision, taken when you name things.
Property columns are never summed. Diameter, height and custom fields marked as not summing describe what a thing is rather than how much of it there is — adding two wall heights produces nothing.
Instead a property is shown only when it is identical on every merged row, and left blank when the merged rows disagree.
That blank is information and it is easy to misread. It does not mean the value is missing from the drawings. It means the rows behind that total were not all the same — 2.4 m and 2.7 m walls merged into one line of wall area. Sometimes that is what you want. Sometimes it is the first sign that a material needs splitting into two, and you have found out before pricing rather than after. The blank tells you they differed; it does not tell you which drawing had which value, so that answer is in the per-drawing detail.
Comments and layers are not carried into the summary. They have nothing sensible to say across eleven drawings, and they survive in the per-drawing detail where they are useful.
Every summary row expands to show its sources: which drawings contributed, how many rows each contributed, and each one's subtotal for the summing columns.
Two columns in the consolidated table are worth reading before the totals: how many drawings a row came from, and how many rows went into it. They are a quick check on a total that looks wrong. A material you expected on six sheets showing one source drawing means five sheets used a different name. A row built from forty source rows where you expected four usually means the material is broader than you meant it to be.
The expansion makes a figure easier to inspect — it does not by itself make it defensible. It will show you a drawing contributing an implausible share of a total. It will not show you that a material appears twice under two names, because those are two separate rows with nothing connecting them, and it will not tell you that two drawings cover the same area.
Property columns are not broken down per drawing at all.
A roll-up that silently skips a sheet is worse than no roll-up, so the run reports three groups: the drawings included and how many rows each produced, the drawings that loaded with an empty material list, and any that failed to load.
Read the empty list every time. A drawing in a takeoff with no items in it might be a sheet nobody has measured yet, a sheet with nothing to measure, or a sheet somebody measured on a different document by mistake — among other possibilities. The roll-up cannot tell you which. It tells you the question exists, which is more than a manual roll-up does.
The summary works within one project. The candidates are the drawings in that project and its folders, with archived folders left out, and you choose which to include.
Only drawings are included — photos, notes and spreadsheets are not material lists and are not counted. Counted items merge exactly as measured ones do, so the fixture counts from counting symbols on a PDF floor plan come through as rows like any other. And archiving matters: superseded drawings moved into an archived folder stop appearing as candidates, which is a practical argument for archiving as you go. Takeoff project templates covers a project structure that supports this.
The roll-up does not reconcile anything. It does not notice that two drawings cover the same area, that one supersedes another, or that a material appears under two spellings. It adds up what you select, which makes the selection a decision rather than a formality — and one nothing else in the process will check for you.
What comes out is quantities. Rates, labour and everything that turns a quantity into a price belong where you price.
Take a job whose total you already know, run the summary across its drawings, and compare it with what you sent out. Where the two disagree, the source counts and the expanded rows are where to start looking — and finding out why, on a job that has already gone, is considerably cheaper than finding out on the next one.
