How do I know when to stop taking one-off custom build jobs?
Short answer
When your R&D and rework hours are consistently eating your margin on novel builds, or when a couple of repeatable builds would fill your bays with far less risk and headache. One-off work is often more prestige than profit once you actually run the numbers on the hours it eats.
One-off work looks impressive and often runs thin on margin
A wild, never-been-done build is great for your shop's reputation and social media, but reputation doesn't pay payroll. If you actually run the numbers on your last several one-off builds, tracking real hours against what was quoted, you may find the effective hourly rate on that work is meaningfully lower than your repeatable jobs, because novel work always eats more hours than estimated no matter how experienced you are.
That doesn't mean stop entirely, it means know the number. A shop that tracks this honestly can decide deliberately how much one-off capacity it wants, rather than backing into it because it's exciting work to take on.
Look at what's filling the bays you're turning away
If you're consistently declining or delaying repeatable, well-margined jobs (common platform swaps, known tune combinations, standard maintenance on your regulars) because a single sprawling one-off project has a bay tied up for months, you're trading predictable profit for a project that might not even finish on budget.
A useful gut check: would you rather have three well-quoted, known jobs filling that same calendar time, or the one prestige build. There's no universally right answer, but most shops answer that question honestly and realize they've been choosing prestige over margin without meaning to.
Set a cap, even an informal one
Decide, even loosely, what percentage of your shop's capacity you want going to one-off or R&D-heavy work versus known, repeatable jobs, and check yourself against it quarterly. Shops that never set this cap tend to drift toward more one-off work over time, because it's the most interesting work to take, not because it's the most profitable.
Comparing one-off build margin to repeatable work
| One-off build: quoted 60 hrs, actual 95 hrs | effective rate drops ~37% below quoted rate |
|---|---|
| Repeatable known build: quoted 40 hrs, actual 42 hrs | effective rate within 5% of quoted rate |
| Bay-months tied up per year on one-off work | track this explicitly, don't estimate |
The math: Track quoted hours vs actual hours by build type over a year. Consistent large overruns on one-off work versus tight accuracy on repeatable work tells you where your real margin lives.
Signs it's time to cap one-off work
- 1.Actual hours on recent one-off builds consistently run 30%+ over what was quoted
- 2.You're turning away or delaying known, well-margined jobs because a one-off project is tying up a bay
- 3.R&D and rework time on novel builds isn't being tracked or billed separately
- 4.You're taking one-off jobs mainly for the social media or reputation value, not the margin
Where owners get this wrong
- Never tracking quoted-vs-actual hours, so you have no real data on one-off profitability.
- Taking every interesting build offered regardless of what it displaces on the calendar.
- Confusing reputation value with actual margin, and letting that confusion drive capacity decisions.
Ask your version of this question
This is one business owner-tested take. Your shop size, trade, and team change the answer. Ask the exact version inside Ask a Shop Owner and get a response grounded in how owners like you actually handled it.
Now run it against your numbers.
This answer is written for shops in general. Inside Ask a Shop Owner it becomes your answer: your revenue, your crew, your market, your history. Ask "How do I know when to stop taking one-off custom build jobs" and get the version that accounts for what you already told us.
Was this helpful?
One click. It tells us which answers to deepen and which questions we still owe you.
Not ready for a trial? Take the scripts with you.
Leave your email and we send three short notes over the next week showing how owners used answers like this one on a real decision. No newsletter, no drip forever.
Three emails, then it stops. Unsubscribe in one click. We never sell your email address.
Answered for these owners