The floor plate kills more conversion deals than the financing does

Your GC is describing something the strategy guide puts plainly: not every building converts well, and office floor plates are one of the primary reasons. The guide specifically calls out floor plate geometry, window placement, and plumbing layouts as the variables that determine whether a conversion can work at all, before financing enters the picture.

Here is what the 62-foot depth number means for anyone reading this who has not held a tape measure in an office building. Residential building codes in most jurisdictions require that habitable rooms, meaning bedrooms and living areas, receive natural light from a window. A typical apartment needs roughly 20 to 30 feet of building depth to make that work with a standard double-loaded corridor (one hallway with apartments on both sides). At 62 feet from glass to core, you are left with a dark interior band that cannot legally or practically become living space. You can chase it with atriums, light wells, or unit configurations that wrap the core, but each solution adds cost and eats rentable area, which is the square footage you can actually charge rent on. When rentable area shrinks and cost climbs, the spread between what the building is worth acquired and what it costs to convert it can disappear entirely.

The thing worth sitting with, and I say this as someone who explains how things work rather than as someone who has underwritten a deal, is that the guide frames feasibility discipline as the core institutional skill here, because construction cost can rival ground-up building on a problem floor plate. A sponsor who prices the acquisition off a rent roll before a conversion architect has walked the building is underwriting a different asset than the one they are buying.

What role are you typically in when you are walking these buildings? That would help me point you toward the part of the guide most useful to where you sit.

The 62-foot depth number is the one I keep coming back to. I did a gut rehab on a converted textile building in Richmond a couple years ago, smaller scale, but even at 18,000 sf per floor we were fighting a 48-foot interior span and I was pulling light wells that cost me $340 a square foot to frame correctly just to get code-minimum egress light into the middle units. And that was a building where the bones were already broken up by old load-bearing party walls that gave us natural unit separations. A clean open plate at 62 feet with a central core is a different animal entirely, because you do not have those accidents of history working for you. The developers I have watched try to solve it with double-loaded corridors end up with interior bedrooms that appraisers just do not know what to do with. Financing was always the story they told the room but the yield-on-cost math was already dead the day somebody drew that core location on a napkin forty years ago.

The thing I have never seen work is underwriting the light problem away with a courtyard cut after the fact. On paper it looks like a solution, you lose maybe 8,000 sf per floor and call the units "loft-style," but the structural cost to open that atrium on a poured concrete building runs $400 to $600 a sf of cut, and by the time you are done you have destroyed the unit count that justified the land basis in the first place.

The deals that actually close on floor plates like that are the ones where somebody with a tape measure killed 40 units in the proforma before the LOI went out, not after.

The point worth pushing back on is treating the floor plate as the kill shot when it is really a sequencing problem. Sponsors walk buildings in the wrong order. They run IRR models and GP decks before anybody has done a daylighting section, and by the time the GC is on site the sunk cost is already doing the work of keeping a bad deal alive. Picture a 140,000 sf former insurance headquarters with 24,000 sf plates, the core offset to one end, and depth from glass on the long axis running 58 feet on most floors. The sponsor knows it. They convince themselves a double loaded corridor with a light well punched at 30 feet is going to pencil, it does not, and they are $180k into predevelopment before the conversation gets honest. That pattern repeats in market after market. The better sequence, and the advice any landlord even sniffing at a conversion play should hear, is to bring whoever is going to frame the interior partitions into the building before you hire the architect for full SD. It need not be a GC, just somebody who has actually built residential in a retrofit condition and knows what 22 usable feet of apartment depth produces in terms of rentable area per floor. You want that number before you build the capital stack rather than after. The floor plate problem is real, but the deeper issue is that adaptive reuse gets underwritten like a ground up deal and it is nothing of the kind. The variables that kill you are baked into the concrete before you show up.

What is the core-to-glass ratio on the floors you are actually watching sponsors skip past?

The part nobody prices correctly is the corridor tax. On a 28,000 sf plate with a central core you are not just losing the dark interior square footage, you are paying to build a double-loaded corridor that serves units on both sides of a hall that already feels institutional the day it opens. We converted a former county office building in Sacramento, 2019 to 2021, plates were 22,000 sf with a 38-foot depth to glass, which is actually manageable, and we still ended up at 78% efficiency after egress and mechanical chases. At 62 feet of depth you are probably looking at 65% before you even start arguing with the fire marshal about egress travel distance, and that missing 13 points is not abstract, it is the difference between 300 units and 245 units on the same debt load.

The thing I have watched fail repeatedly is the workaround of amenitizing the dark core, turning it into a coworking lounge or a fitness room or some kind of programmed social space, with the argument that it offsets the unit count hit by justifying higher rents. Does not work. Renters in the East Bay submarkets I track are paying for light and for ceiling height, and a curated dark lounge does not move the needle on a lease comp. What happens instead is you spend $800,000 to $1.2 million building out a core amenity floor, you still have 245 leaseable units, and your per-unit cost basis is now $15,000 to $20,000 higher than a comparable new construction deal that never had the floor plate problem to begin with.