Most architects leave a firm because they want control over the work. The design, the clients, the way a project actually gets run. Very few leave because they were excited about cash flow forecasting, and that is usually where the first hard year comes from.
The design side tends to take care of itself. You already know how to do that. What catches people out is the eighteen months between signing an office lease and having a practice that reliably pays for itself. Preparing for that stretch is almost entirely financial, and most of it can be done while you’re still employed.
Here’s what’s worth sorting out before you hand in your notice.
Work out how long you can go without a paycheck
Start with the number. Not the firm’s budget, your household’s. Rent or mortgage, insurance, groceries, student loans, everything that has to be paid whether or not a client signs anything this month.
Multiply that by twelve, then be honest about whether eighteen is closer to the truth.
Architecture projects are slow to start and slower to pay. A contract signed in March might not produce a deposit until May and a second invoice until August. Practices that close early usually do it because the founder ran out of personal runway, not because the work wasn’t good.
Practices that close early usually do it because the founder ran out of personal runway, not because the work wasn’t good.
Building that cushion takes time, which is exactly why it belongs to the year before you leave rather than the month after.
Treat your personal credit as the firm’s credit
In the beginning, there is no separation between the two. A new practice has no financial history of its own, so a bank reviewing a line-of-credit application, an equipment lease, or a personal guarantee on office space is really looking at you. Your score sets the terms you get offered.
That’s worth checking long before you need it, and credit monitoring makes it simple since banks and card issuers now include it for free. It shows you where you stand and tells you when something on the report moves, which matters because reporting errors are common and take weeks to correct. Catching one eight months before a financing application is paperwork. Catching it the same week is a problem.
Catching one eight months before a financing application is paperwork. Catching it the same week is a problem.
It also helps to avoid opening or closing accounts in the months before you apply for anything. Lenders like stability, and a file that has just been rearranged tends to get priced accordingly.
Separate the money before the first invoice
Set up the entity and the business account before you take on any work at all. Running early projects through a personal account is tempting when there’s only one client, and it creates a mess that a bookkeeper will later charge you to untangle.
The structure itself is a conversation for an accountant and, given the liability exposure in this profession, an attorney. Whatever they recommend, open a dedicated business checking account, get a business card, and connect both to accounting software on day one.
Reconstructing a year of mixed personal and project transactions at tax time is a miserable way to spend a January.
Plan for the gap between doing the work and getting paid
Design fees arrive in stages, and those stages rarely line up with your own expenses. Payroll, rent, and software subscriptions run every month without fail. A schematic design phase might take ten weeks and bill once at the end of it.
Structure the agreements to close that gap wherever you can. A retainer at signing, invoices tied to phase completion instead of calendar quarters, and clear late-payment terms all help. So does opening a modest line of credit while you still have employment income on the application, because it is far easier to arrange before you need it than during the month you do.
Budget for the costs that don’t announce themselves
Professional liability insurance is the one that surprises people. It’s a serious annual cost, it’s often required by contract, and it generally needs to stay in force long after a project wraps up.
Then there’s general liability, workers’ compensation once you hire, licensure renewals in every state you practice in, continuing education, professional memberships, and the software.
A single seat of Revit or Archicad is real money, and it recurs whether or not the month was busy. Write it all out with actual figures before you commit to anything. The total is almost always higher than the version people carry around in their heads.
Decide how you’ll charge before a client asks
Underpricing early is common and very hard to undo. The fee you accept on the first project becomes the reference point for the second, and clients in the same market talk to each other.
Work out your true hourly cost, including overhead, the benefits you’re now paying for yourself, and unbillable time, which takes up a much larger share of the week than most people expect. Then decide where a percentage of construction cost makes sense, where a fixed fee does, and where hourly is the honest answer. Having that settled in advance keeps a scope conversation from turning into an improvised negotiation.
Leave in stages if your situation allows it
Not everyone can, and employment agreements vary quite a bit, so read yours closely before doing anything. Where it’s permitted, though, taking on a small project or two alongside your existing role is the closest thing to a test run you’ll get.
It shows you what the administrative load really feels like, whether clients come to you without a known firm name behind it, and how long the payment cycle actually stretches. It also builds the cushion. Leaving with one signed contract and a few extra months of savings is a very different start than leaving with a business card and a plan.
Start earlier than feels necessary
None of this requires a polished business plan. It requires a savings target, a clean credit file, an entity, a separate account, and a realistic view of what the year will cost.
The practices that make it through the first two years are rarely the ones with the strongest portfolio at launch. They’re the ones that could afford to wait for the right work instead of taking whatever turned up first.
Most of it can be handled quietly over twelve months while a salary is still coming in. The practices that make it through the first two years are rarely the ones with the strongest portfolio at launch. They’re the ones that could afford to wait for the right work instead of taking whatever turned up first.
