The question is whether a 12 unit project in a mid-size Pacific Northwest market can carry its soft costs at all, or whether the math only starts working somewhere around 30 or 40 units.
Twelve can work, but soft costs per door have to stay under roughly $18k to $20k or the deal bleeds out before ground breaks. The number to pin down first is what the local jurisdiction charges in system development charges per unit, because in some PNW markets those alone run $15k or more, and that is before a single architect invoice.
Land basis per door is what usually kills a 12 unit project, long before the soft costs get a chance to.
Take a 16 unit ground-up in Tacoma in 2021 where the developer already owns the dirt. The math works, but only barely. Soft costs run about 28k per door all in, architect, geotech, civil, permits, the whole pile, and on a 12 unit that same number is a reason to walk. What saves a project like that is a construction loan at a spread that does not destroy the yield on cost, and finding it usually means going through three lenders. The 30 to 40 number people throw around holds up reasonably well, and the reason is that below 20 doors you have almost no room for a cost overrun before the deal is underwater on a refi or sale. If you are not controlling land basis, 12 units in a mid-size PNW market is a very hard ask.
Twelve pencils once the model stops assuming rents hit pro forma by month three.
What's your construction cost per square foot coming in at, and are you doing podium or stick-on-grade?
That spread between those two methods can swing your per-door cost by 40k or more, which at 12 units is basically the whole question.
Twelve works fine in Bend or Boise right now. Skip Tacoma.
Modular or panelized framing can cut soft cost per door at 14 units by about 18 percent.
