Almost every kind of retail rewards size, which is why the small architectural salvage dealer shouldn't survive, and does. Buy in bulk, standardize, undercut on price, and the big operation wins. Salvage is one of the few trades where that math fails, and the reason is sitting on the bench: one-off parts, odd fittings, and work that changes piece by piece.
One-off inventory, hands-on work
Scale depends on selling the same item a thousand times: reorder, restock, forecast demand. Salvage inventory is unrepeatable. Every piece is a one-off. You can't reorder a particular hundred-year-old fixture; when it sells it's gone, and the next thing through the door is different. There's nothing to standardize, no supply chain to tune, no catalog that stays put. The machinery that makes big retail efficient has nothing to grip.
The work resists automation just as stubbornly. Restoring a fixture is judgment applied one piece at a time: seeing what it needs, rewiring it safely, deciding how far to take the finish. It won't run down an assembly line, because every piece arrives in a different state and asks a different question. The knowledge it takes, what a part is, what era it belongs to, whether it can be saved, lives in a person holding the piece, not a process.
This is where the small dealer wins by default. One person who knows the material deeply can judge each piece on its own terms, restore it properly, describe it honestly, and answer the odd question a buyer actually has. That depth is the product. Scale would only thin it.
So salvage stays a bench trade almost by its nature. The pieces are singular, the work is hands-on, and the price depends on one person's judgment instead of a system's efficiency. A warehouse can't stock a one-off fixture, restored and understood by the person who pulled it out of the wall. That part only comes from a shop small enough to know it.