Everyone tells beginners to start with a small shell. Is a 6,000 sf retail building actually easier, or just smaller?

The small deal is harder to fund, which tells you something. A $2m shell doesn't clear the fixed cost of underwriting for most institutional money, so you're at a small bank or friends and family, and those sources ask for personal guarantees. The larger deal has more sophisticated money in it, which means more scrutiny and also more people checking your assumptions before you commit.

harrow, your soft cost point is the one I hadn't sized properly. Do those actually run flat? I assumed civil scaled somewhat with site area at least.

Some of it scales with site area, kestrel, and some of it doesn't move at all. Geotech scales with the number of borings, which follows building footprint loosely. Civil engineering fees follow site complexity, so stormwater on a 1.5 acre pad with an infiltration requirement can cost close to what a 4 acre site costs. Traffic studies, environmental phase one, ALTA survey, utility application and impact fees, and the legal work on entity formation and the lease form have a floor that a smaller building does not lower. Impact and connection fee schedules vary by jurisdiction and some are calculated per fixture unit or per trip rather than per square foot, so check the actual local schedule before you assume.

The practical result is that soft costs plus site work often run a higher percentage of total project cost on a small shell than on a large one. That's not an argument against starting small. It means the yield on cost has to be underwritten with those numbers in hand, and the land basis has to be low enough to carry them.

On harrow's actual question, the two paths teach different things. Building a small shell teaches entitlement, draw management, and dealing with a municipality, which is knowledge you can't buy. Writing a check into someone else's project teaches you how deals are structured and what a capital call feels like, and it teaches nothing about running a job.

What decides it is whether you can lose the money and keep operating. Both paths have a version where the outcome is zero, and passive positions in development are illiquid for years.

Do the small one but pick a use where the tenant exists before you break ground. Spec retail in a market with three prospects is how people learn this lesson expensively. I've watched a 7,000 sf shell sit at 40 percent leased for 14 months while the loan converted and the sponsor covered the gap out of pocket.

Dumb question, when quarry says the loan converted, converted to what? I thought a construction loan just gets paid off when you sell.

@compass a construction loan is short term and interest only during the build, and at completion you either sell, refinance into a permanent loan, or the loan has a built-in conversion to a term loan on stated conditions. Those conditions usually include a minimum occupancy or a debt service coverage test. If the building isn't leased enough to hit the test, the borrower has to pay down principal or find other financing. Terms differ by lender, so anybody doing this should get the conversion conditions in writing up front.

A few deals in, what surprised me is how much of the work is municipal and how little is construction. I spent nine months on approvals for something that took five months to build. If you hate meetings where nobody decides anything, small ground-up is going to be miserable for you at any size.

Voted for the equity check. I'd rather pay tuition to watch someone competent do it than pay tuition to my own mistakes on a guarantee.

quill, the problem with tuition-by-LP-check is you don't see the parts that matter. You get a quarterly update. You never see the sponsor argue with the fire marshal about a second exit, and that argument is where the money goes. I'd rather be small and in the room.

harrow, the third option is neither. Buy the land, entitle it, sell it entitled to somebody who builds these for a living. You learn the front half of the process, which is where most of the value creation sits anyway, and you never touch a construction loan or a general contractor. Timeline risk is still yours and entitlement can fail outright, so it isn't safe. It's just a different risk than building.

Coming from the construction side I'll say the shell is genuinely simple to build and genuinely not simple to own. The trades on a 6,000 sf pad building are straightforward and a competent local GC does them in their sleep. What'll get you is tenant improvement allowances, because you'll sign a lease promising a finish-out you priced from a spreadsheet and the actual buildout comes back 30 percent higher. That number belongs in your underwriting with a real contingency on it.

girder's TI point is what I keep seeing in the deals I look at. The shell number is defensible and the allowance is a guess. Also worth saying that a lot of the healthcare and clinic-shape buildings around me pencil better than retail right now, same construction, tenant with a longer horizon.

Small and boring, with a tenant signed first. Every deal I've done that went badly went badly because I built for demand I assumed instead of demand that had signed something.