How we decide whether a deal is actually good.

Start with the exit. Then attack every assumption between “cheap” and “profit.”

1. Start with the real exit value

Retail price is mostly trivia. We want realistic completed-sale value, local demand, time to sell and what the item moves for when you want the cash sooner rather than later.

2. Identify exactly what you are looking at

Model number, configuration, battery, missing pieces, material, fitment and condition can move value by hundreds or thousands of dollars. “Looks like one online for $2,000” is not research.

3. Build the conservative math

Subtract fees, repairs, consumables, transport, storage, cleanup and a risk allowance. The maximum buy price comes from the exit backward—not from negotiating ten percent off the seller's fantasy number.

4. Add the physical-world problems

Can you lift it? Load it? Power it? Store it? Get the seized bolt out? Dispose of what remains? A spreadsheet profit that requires a forklift you do not have is not a profit yet.

5. Find the fallback exits

Complete resale is plan A. Repair, trade, part-out and scrap may be plans B through E. A good asymmetric deal often has a surprisingly solid floor even when the best-case sale fails.

6. Prefer evidence over seller stories

Manufacturer documents, completed sales, actual material prices and independent repair experience beat asking prices and “it worked last time I used it.” When evidence is weak, we say so.

7. AI-assisted, human-reviewed

AI is used to search broadly, compare sources, follow references and pressure-test the math. Important claims still get checked against the underlying evidence. The point is deeper investigation, not mass-producing thin articles.