Redlines are not a punishment. They are the test, and almost nobody tells you that on day one. A practice hands a new starter a marked-up set because it is the cheapest possible way to find out three things: whether you can follow a convention you did not invent, whether your work comes back needing a second check, and whether you understand what the drawing is actually for. Pass that test quickly and you move. Fail it slowly — by producing work that is technically correct but off-standard, or by waiting to be asked — and you can spend eighteen months amending bathroom layouts while someone six months ahead of you is running a planning submission.
So the goal of the first ninety days is not to be liked. It is to become expensive to waste. Here is how that breaks into three phases, and the specific behaviours that shorten each one.
Phase One, Weeks One to Three: Become Fluent in the House Style
Every practice has a dialect. Layer or class naming, line weights, which pen set plots what, how a door is tagged, where the north point sits, how sheets are numbered, what goes in the title block revision column and what does not. None of this is in your degree. All of it is what the person checking your work sees first.
On day one or two, ask for the office standards document. Those exact words. Some practices have a tidy BIM Execution Plan or CAD manual; plenty have a half-finished Word document from 2019 that only one associate knows about; a few have nothing but a template file and folklore. All three answers are useful. If it exists, read it properly in your own time and keep it open while you work. If it does not exist, say — politely, once — that you will write down the conventions as you learn them, and then actually do it. In practice, the new starter who turns up at week three with a one-page crib sheet of the office's own standards becomes the person who gets asked to set up the next project's sheets. That is not a hypothetical. It is one of the fastest promotions from redlines I have seen.
Second thing to find in week one: the file logic. Where does the live model sit, where do superseded issues go, what is the naming protocol for an issued PDF, which folder do consultant drawings land in and who downloads them. Saving a file in the wrong place is the kind of error that does not get you shouted at but does get you quietly reclassified as someone who needs supervising.
Third: read one complete, archived package end to end. Pick a project that got built, find the construction issue set, and read it like a book — general arrangement plans, then sections, then the details they reference, then the schedules. You will learn more about how this office thinks in an afternoon than in three months of being drip-fed tasks. Ask your line manager which project to read. They will tell you, and they will remember that you asked.
Learn who checks your work — and read the redlines as a document
Find out, specifically, whose pen is on your drawings. It is often not your line manager. In a lot of studios the person who checks is a mid-level architect or an associate with a reputation for catching things, and their markups are a free map of what this practice cares about.
So keep them. Before you make the amendments, save the marked-up PDF somewhere of your own. After the third or fourth set, go through them together and look for the repeat offenders. Is it always dimensions to structure rather than finish? Always the same hatch? Always a missing reference on a section mark? Those repeats are your personal standards document, and when they stop appearing, the checker notices before you do.
Phase Two, Weeks Four to Eight: Get One Small Thing That Is Yours
The jump off redlines is not a jump onto "design". It is a jump onto owned scope — one discrete element of a real project that is your responsibility to resolve, draw and keep coordinated. A bin and bike store. A WC core. The balustrade detail. A door schedule. The window schedule for one elevation. Small, boring, finite, and absolutely yours.
Owned scope is the unit of trust in an architecture practice. It is how a project architect assesses you without risking anything, because if your bike store is wrong, the damage is contained and fixable. Once you have resolved one piece properly — drawn it, checked the clearances, found the manufacturer's data, coordinated it with the drainage, issued it, and had it come back with almost nothing on it — the next thing you are handed will be bigger. That is the whole mechanism. There is no other one.
So ask for it, and ask correctly. Do not ask for "more design work", which lands as a complaint about the work you have. Ask for a specific thing, at a moment when the team has a gap, and attach a plan. Something like:
"I've finished the amendments on the ground floor set. I noticed the refuse store hasn't been resolved yet — would it help if I took that away, checked it against the council's storage requirements and came back with two options for you to look at on Thursday?"
That script works because it does three things at once: it reports completion, it names a real gap in the package, and it offers a deadline the other person did not have to set. It also gives them an easy no. If the answer is no, ask what would be useful instead, and take that.
If you are in a small practice, this conversation is usually with a director, and it is faster and more informal — small offices hand out scope early because they have to. In a large practice the gatekeeper is the project architect or job runner, not the director, and going over their head to ask a partner for design work is the single most effective way to make yourself unpopular in month two. Work out which building you are in before you start knocking on doors.
What you are actually being judged on in a desk review
When you bring that first piece of owned scope back, you are not being judged on whether it is beautiful. You are being judged on whether you can explain why it is the way it is, in order, without waffle, and whether you brought options rather than a single fixed idea you are emotionally attached to.
This is the same discipline as a crit, and it is the bit of architecture school that transfers most directly into practice: a clear narrative, three or four decisions that follow from it, and the self-criticism to say "this part isn't working yet, here's why" before someone else says it for you. In the mentoring sessions we recorded for the Architecture Student Guide, the strongest moments came exactly when the student said out loud that the plan felt too boxed-up and the circulation was dead — naming your own weak point is a senior person's habit, and doing it in month two is disproportionately impressive.
Phase Three, Weeks Nine to Thirteen: Be Useful on Something Live
The last phase is where trust stops being about competence and starts being about judgement. Competence is drawing what you were asked to draw. Judgement is noticing the thing nobody asked you to look at.
The highest-value version of this in practice is the coordination catch. You have the consultant drawings in the project folder; almost nobody junior ever opens them. Overlay the structural grid on your plan. Underlay the services layout. Check that the duct running down the corridor still fits once the ceiling void is as shallow as the section says. Check the stair the engineer has drawn is the stair you have drawn. When you find a clash — and you will — do not announce it to the room. Send it to the project architect, in writing, as a question:
"The M&E layout issued Tuesday shows a 400mm duct in the corridor ceiling, but our section 02 gives a 300mm void at that point. Have I read that wrong, or should I flag it for the next coordination call?"
Question, not accusation. Written, so it exists. Specific drawing references, so it can be checked in thirty seconds. Do that twice and you become the person whose eyes are worth having on a set.
The same applies to planning and regulatory risk. You do not need to be a Building Regulations expert in month three, but you should know which Approved Documents bear on the project you are on and be able to say "this escape distance looks long, should I check it against Part B?" rather than silently drawing it. Knowing where the question lives is most of the skill. Nobody expects you to know the answer.
The Admin You Must Not Leave Until Month Six
Start your PEDR in week one. The ARB's criteria require a recorded period of practical experience before you can sit Part 3, and the RIBA's PEDR is the standard way of logging it with an employment mentor signing it off. Nobody will chase you for this. Everyone who leaves it until the end regrets it, because by then you are reconstructing nine months of work from memory and your timesheets.
Log it properly: not "worked on housing scheme", but which RIBA Plan of Work stage, what you personally did, and what you learned. That record is also the thing you take into your probation review, your appraisal and your next interview. A new starter who can say "I've had Stage 3 and Stage 4 exposure on two projects and here's what I did on each" is a different candidate from one who says they were in the office for a year.
And while you are at it: find out when your probation review actually is, and what is being assessed. Ask in week two. The answer takes thirty seconds and the question signals that you intend to pass.
Ninety Days Is Not a Deadline, It Is a Gradient
None of this is about hustle. It is about removing the three reasons a practice has for keeping you on redlines: your work comes back needing fixing, nobody knows what you are good at, and you have never been tested on something that mattered. Close those off in order, and the work changes shape on its own.
If you want the longer view of what the first couple of years actually feel like, we've written about life after graduating elsewhere. But for the next three months, the whole job is this: learn the dialect, own one small thing completely, then open the drawings nobody asked you to open.
