A spec build case where using an owner's own crews turned a thin win into a real loss

The labor accounting is real, thatch, but the bigger dollar sits in the schedule. Fifteen months on a 2,350 foot single family with an existing plan and a platted lot is a long time, and internal crews are the usual reason. Your techs had a boss with paying customers, and that boss was you, so every time a service call came in the house lost the day. A sub with a signed scope and a retainage schedule does not get pulled, or if he does you get to hold money back. Contractors who run spec houses with an in-house framing and trim crew tend to report the same pattern. The framing is always fine. Everything after framing turns into a queue behind whatever is more urgent that week, and the queue is what costs you.

The same shape shows up at a much smaller scale. Say an owner puts a maintenance person already on payroll onto four unit turns in a nine unit building, reasoning that the labor is already paid for. The turns run six weeks each instead of two, and two units sit vacant and bleeding while punch list work drags on. Priced against what a turn crew actually charges, the savings run around 1,400 dollars while the vacancy cost runs over 9k. The lesson holds broadly: a salaried person is not free labor, they're already spoken for.

What did the 88k lot actually carry in terms of time? Platted and permit ready are different animals in most jurisdictions, and the difference lives in things like utility connection scheduling and whether the county wanted anything on drainage. Fifteen months on the build side sometimes has three or four months of that being a lot problem, and it gets assigned to construction because that's when the calendar started running.

A little pushback here. The internal labor point is an accounting insight and it's a good one, but that isn't the loss. Take a build where hard cost comes in 13 percent over a budget on a first ground-up project, that's honestly survivable. The extra six months on the schedule is the loss. Carry, taxes, insurance, and the risk that the comp set moves underneath the project while it's standing half finished. It's common to see two builds on the same operator's track record where the exit price at month nine and the exit price at month sixteen differ by more than the entire profit. A deal that pencils to a thin win on paper often owes more to a market holding steady than to the underwriting.

It depends on the loan structure. Some construction loans carry an interest reserve in the budget that gets drawn down as draws happen, so the carry is invisible until the reserve runs dry. Past that point, or on loans with no reserve, the borrower writes monthly interest checks out of pocket. A 34k carry on a build that ran 15 months against a 9 month underwrite usually reflects reserve depletion followed by months of direct payments.

Thatch, when you say price your own labor at market, what number do you actually put in the budget line? Your billed rate includes your overhead and your profit, so if you use it you're charging the project for profit you're also collecting as the builder. Do you use full billed rate, or loaded cost plus something? I'm building my first budget spreadsheet and this is exactly the cell I keep staring at.

Honest read, at month 6 with the house dried in and your service business bleeding accounts, I'd have listed the shell to another builder and eaten a small haircut. Thin margins are my whole life and the one thing I've learned is that a project that is destroying a different business you own has already cost more than it can pay you. You held it 9 more months to protect 18k.

Did the service company invoice the project at all, even at cost? Asking because I've got six months of research and zero deals, and the thing I'm noticing is that the people who lose money on paperwork usually lost it because two entities they own were doing business with each other on a handshake. Sounds like the money moved but the invoices didn't.

The sensitivity is what I'd have wanted on paper before the lot closed. Build the budget at 9 months and also at 15, with carry, and also at a 10 percent hard cost overrun, and see what the worst intersection looks like. Thatch's worst intersection was roughly break-even before you even count the service business, and that's knowable in advance with a spreadsheet and an afternoon. The internal labor discount just moved the whole grid up one notch and made the bad cases look fine.

In markets where 628k is not a house but four houses, this reads differently. In a 90 to 180k price band, a 53k swing is the entire project. Does not change the lesson, just means the cushion available to absorb it does not exist everywhere.

Good trades hate being the in-house crew on the boss's side project. There's no scope, no change order, and no end date, and they know the service calls are what actually pays them. I've watched two shops lose a lead tech to a spec house that dragged. Ask your techs what those months were like, thatch, you might find the 71k has a tail on it.

Better to never combine the two. Boring, but it works. A service business should get all of the operator's attention, and someone who wants a house is better off as a customer of a builder who signs a fixed scope.

How would you have known in advance which accounts were at risk? I'm trying to work out where I fit and this thread makes it sound like the service side has this hidden fragility where losing two contracts is what breaks the year.

The line worth writing down is that if a deal only works because the crew works for free, it doesn't work.

Different angle, thatch. If you'd taken that same summer and managed someone else's spec build for a fee, with your crews staying on their routes, what would that have been worth? I keep looking at the service side of development as the safer seat and this thread is the first time I've seen the actual comparison sitting in one place.