The building has to pass an inspection. The drawings that describe it never have to pass anything.
I have sat through plan sets from homebuilders, school districts, hospitals, and city halls, and I have never once seen two architecture firms lay a set out the same way. Not the sheet numbers, not what gets bookmarked, not how a room gets tagged, not even how a plain fact like a ceiling height gets written down.
Everybody who has spent real time on the receiving end of these documents already knows this. What I never had a good answer for is why it happens, what it actually costs, and who ends up paying.
The permit checks the building, not the drawing set
Construction runs on standards for almost everything that gets built, and the permit is what puts teeth in them. A government office checks the design against the code before anyone pours a foundation.
Nothing in that process, or anywhere after it, checks how the drawing set itself is put together.
A standard for that does exist. The US National CAD Standard is a consensus document combining the American Institute of Architects' CAD Layer Guidelines, the Construction Specifications Institute's Uniform Drawing System, and the National Institute of Building Sciences' plotting guidelines.
I could not find anything that requires a firm to use it, or that checks whether they did.
One buyer big enough to make it stick
There is exactly one place I could find where this gets enforced, and it explains the whole pattern.
The General Services Administration requires the architects and engineers preparing CAD deliverables for it to comply with the national standard, on top of its own additional CAD requirements, and states plainly that it will not waive either one.
GSA can do this because it is the same buyer over and over, ordering enough buildings that it is worth the friction of making every architect who works for it draw things its way.
Almost nobody else hiring an architect is in that position. A hospital system doing one expansion. A city building one new hall. A homebuilder working with a firm for the first time. None of them have the repeat volume or the leverage to make demanding a format worth the fight.
Nobody is being lazy. Nobody with enough power over most individual jobs has ever had a reason to fix it.
House style is not an accident
There is a real reason firms keep their own way of doing this, and it gets missed every time this comes up as a complaint.
How a firm lays out its sheets, numbers its rooms, and organizes its set is closer to a house style than an accident. It gets built over years by whoever runs the office, shaped by the software the firm came up on and by the kind of buildings they actually specialize in.
That is part of how a firm works, not incidental to it.
This is where I think a lot of construction AI gets it backwards. A tool that quietly expects every firm to reformat its habits to match what the software wants is not asking someone to learn a new button. It is asking a firm to give up a piece of how it works in order to be read by a machine.
That is not stubbornness. That is a real cost, and it is a reasonable thing to push back on.
Everyone who touches the set pays an entry fee
Here is the part I underrated when I first wrote about this. The variance does not only cost the estimator reading the set. It charges an entry fee to every working relationship the firm has.
| Who pays it | What they have to relearn | When it repeats |
|---|---|---|
| A new hire or intern | Sheet organization, tagging, templates, naming, and everything the office never writes down | Every move between firms |
| A collaborating firm | How the other practice draws, before coordination can start | Every new partner on a job |
| An outsourced drafting shop | A separate set of conventions for every client practice it serves | Every client it takes on |
| The estimator and every trade | Where anything is, before pricing can begin | Every bid, won or lost |
An architect or an intern moving from one shop to another brings the general concepts with them. They still have to learn how that particular office sets things up: the sheet organization, the tagging, the templates, the naming, and everything the office considers so obvious it never writes it down. None of that transfers.
Two firms collaborating on the same job have the same problem pointed at each other. Each one has to learn how the other draws before the coordination work can even start.
The clearest case is the drafting shops. Firms that outsource production, offshore or domestic, work for many architecture practices at once, which means holding a separate set of conventions for every client.
The vendors who sell that work describe it openly. Their pitch is that drafters are hired for one firm, trained on its standards and workflows, and assigned to it long term, working inside that firm's own template, view templates, and family library. At least one recommends the National CAD Standard specifically as the way to keep drawings consistent across the arrangement.
Read that from the other direction. The reason a drafting shop has to sell dedicated, firm trained staff is that the generic alternative does not work. There is very little competence that transfers between firms. There is per firm competence, and it has to be built.
And it is not paid once. Every time the staff changes, the software changes, the office standard gets revised, or the firm picks up a new partner, the same learning happens again.
The cost lands on the architect too
The easy read is that the architect sets the conventions and everyone downstream pays for them. That is only half of it.
The variance makes working with any given firm more expensive to start. The first job with a new practice carries a cost the fifth one does not, and that cost is real whether or not anybody writes it on an invoice.
It also cuts the other way, which is the part worth sitting with. Once a client, a consultant, or a drafting shop has absorbed one firm's conventions, moving to a different firm means paying that entry fee again from zero. The absence of a standard quietly raises the cost of leaving.
That is a switching cost, and it protects the incumbent. It is also a tax on every new relationship the firm tries to start.
The same thing that makes a client hard to lose makes a client expensive to win.
I do not think firms designed it that way. I think it is what happens when nobody with leverage ever asked for a format, and the result is an industry where the cost of learning to read a set gets paid over and over, by everyone, including the people who drew it.
The party best placed to fix it has the least room to
There is an obvious question sitting under all of this. If the format is the problem, why doesn't the profession that sets it just agree on one?
Because agreeing on one is real work. Rebuilding templates, retraining staff, reissuing the office standard, and absorbing the slower months while everybody adjusts. None of it is billable to a project.
And it lands on the party holding the smallest slice of the budget. Design services are commonly quoted somewhere around five percent of total project cost. I have not been able to source that number properly and would treat it as a rule of thumb, but the rough proportion is the point. The party with that share is the one whose output every other party has to read, and the one carrying professional liability for what it says long after the building is finished.
Everyone downstream would benefit from a standard. None of them are paying for it.
That is not a character flaw in the profession. It is the incentive structure working exactly as built.
There is a longer piece to write about what architects carry relative to what they are paid, and I intend to write it. This is not that piece.
What I could not confirm
A few things I went looking for and did not find, which are worth stating rather than hiding.
- I could not find a study that prices the relearning. Not for a new hire, not for a consultant, not for a drafting shop taking on a new client. The outsourcing evidence above comes from vendors describing their own service. That tells you the onboarding is real and routine. It is not an independent measurement of what it costs.
- I could not find a turnover figure for architecture staff that I would repeat. The AIA publishes compensation and benefits data, and the summaries describe turnover as relatively low in 2024, but I have not verified a number I would put in front of anyone.
- The five percent design fee is a rule of thumb I have heard used for years and have not been able to source properly. It is in this piece because the rough proportion matters to the argument, not because I can defend the exact number.
- On the industry's formal fix for an unclear drawing, the clearest measurement I found is a 2013 analysis by the Navigant Construction Forum putting the median reply to a request for information at 9.7 days, rising toward 10 days on projects running two years or longer. That figure comes from one project collaboration platform's own database, drawn mostly from projects in Australia and New Zealand. It is evidence that the formal channel can run slow. It is not a verified number for a typical American job.
- What happens earlier, at the bid, where the deadline is tighter and does not move for anyone's answer, I have not found studied at all. I am arguing that one from what I have watched.
Why this shapes what I build
I am not trying to talk any firm out of its house style. I do not expect the industry to get a mandatory drawing standard, and I am not convinced it should.
What I do think is that the software should carry the cost of the difference instead of pushing it back onto the firm. That is why the document reading side of Setmark starts from what a set says about itself, its own legend, its own schedule, its own conventions, rather than assuming the last firm's habits apply to the next one.
A person still checks what it reads. That is the point of it, not a way around it.