How many criteria should a buy box have?

Six or seven: geography, price band, return floor, property type, and two or three physical minimums. Past that you're usually encoding preferences rather than requirements, and preferences belong in the decision, not the filter.

Should the return floor be a hard rule?

Treat it as hard for the first year. Writing a number down is how you stop negotiating with yourself at 11pm about a property you've already fallen for. Revise it deliberately between deals.

What if my buy box returns nothing?

Relax one criterion at a time and note what each one buys you. Usually the binding constraint is the return floor against the price band, which means your target doesn't exist at that price in that market. Widening the geography often costs less than dropping the floor a point.

How often should I revise it?

After every property you pass on, note which criterion did the rejecting. Three passes in a row on the same criterion means it's either correctly binding or wrong, and only the pattern tells you which. Rates and insurance move fast enough that an annual review is a floor.

How to build a real estate buy box

A buy box fails in one of two ways. It returns nothing, or it returns everything.

Both come from building it in the wrong order: picking a return target first, then reverse-engineering a price band and a geography that could theoretically produce it. What you get is a box describing a property that doesn’t exist.

Build it from your constraints instead. They’re already fixed.

Start with geography

Every other number is downstream of where you’re buying.

A 7% cap rate is unremarkable in one metro and doesn’t exist in another. Insurance ranges from a rounding error to the second-largest line on the operating statement depending on wind exposure. Effective property tax rates vary by roughly seven times across states, from under 0.3% in Hawaii to well over 2% in New Jersey.

Pick one metro. Inside it, name the submarkets you’ll buy in and the ones you won’t, and write down why for each exclusion. You’ll want those reasons in four months, when something in an excluded area looks good on paper and you can’t remember what you knew.

One metro is also a practical ceiling. Knowing a market well enough to spot a bad block takes months, and an investor spread across four metros knows none of them.

Derive the price band from cash

Take what you can put down. Subtract closing costs, which run 2% to 5% of the purchase price depending on your points and prepaid escrows, and land higher on an investment loan than on a primary residence. Subtract the reserve you won’t touch. Subtract whatever the property needs in its first ninety days.

What’s left supports a purchase price at your down payment percentage.

The top of the band is not the most expensive house you can close on. It’s the most expensive house you can close on and still hold through a four-month vacancy and a roof.

The bottom of the band matters too, and gets skipped. Below a certain price in most markets you’re buying a different asset class, with different tenants, different turn costs and thinner financing options. Set a floor and make yourself justify going under it.

One return floor, with the metric named

This is where most buy boxes go soft. “Good cash flow” rejects nothing. “8% cash-on-cash” rejects a great deal at current rates, which is at least honest.

A cap rate target and a cash-on-cash target at the same number describe entirely different properties, because one ignores your financing and the other is mostly about it.

If you care about	Use	Because
Monthly cash in hand	Cash-on-cash	It’s the only one that includes your debt
The asset independent of financing	Cap rate	Comparable across buyers with different loans
A Fix & Flip or a BRRRR exit	Margin against ARV	The rent is temporary

Set the number where it rejects most of the market and admits some of it. If your floor admits nothing inside your price band, the two are incompatible and one of them has to move.

Physical minimums come last

Beds, baths, square footage, year built, lot size. They’re cheap to apply and they feel productive, which is exactly why they creep to the front of the box and start doing work they shouldn’t.

A three-bed minimum is a reasonable filter. A three-bed, two-bath, post-1980, garage, no-HOA filter is a description of a specific house you already have in mind.

A year-built floor is a proxy for wiring, plumbing or roof age. Filter on those directly where the data exists; a maintained 1958 house and a neglected 1994 one are not what the year suggests.

Calibrate against live inventory

Run the box against current listings and count what comes back. The count tells you what you’ve built:

Zero. Your criteria describe a property this market isn’t currently offering.
Two hundred and forty. You won’t underwrite 240 properties, so you’ll choose on gut and the box did nothing.
Fifteen to forty. A list you can work through before the good ones go.

There’s no correct number, since it depends on how much underwriting you can absorb in a month. But you’ll feel the difference immediately between a filter that leaves you with a list and one that leaves you with a market.

When a box returns nothing, relax one criterion at a time and note what each one buys you.

Running the box across a whole metro rather than the first page of listings is the part that used to be impractical by hand. CapScout sends the price band and the physical minimums to the listings provider itself and reports how many homes in the market sit inside that box. Your return floor isn’t in that number and can’t be, because it needs a per-home appraisal, so read the count as the inventory to underwrite rather than the shortlist.

Revising it

Between deals, never during one.

Note which criterion rejected each pass. A dozen passes in, one of them will be doing most of the work, and that’s the criterion worth arguing with.

General information, not investment, tax or legal advice. Rates, insurance pricing and assessment practice vary by market; confirm figures for your own.

How to build a buy box for real estate investing