Pull three comps in almost any established submarket and you will find a pattern that trips up spreadsheet-first investors: the largest home frequently posts the lowest price per square foot, while a smaller home on a better street posts the highest. Most people see that and assume bad data. They throw out the number that doesn't fit, average the rest, and move on. They just deleted the most useful piece of information on the page. Size dilutes. Location concentrates. When you understand why that happens, you stop guessing at value and start reading it.
I have made most of my money in single-family real estate by respecting that one relationship. Buy the cheapest house on the best street, renovate it exactly to the standard that street already proves it will pay for, and let the location do the heavy lifting on the exit. It sounds almost too simple. The discipline is in refusing to talk yourself out of it when a bigger lot or a taller ceiling starts whispering that more square footage is more money. It usually isn't.
Start with what price per foot actually is. It's a ratio: sale price divided by finished square footage. People treat it like a fixed rate, a per-unit cost you can multiply out the way you'd price lumber. That's the error. It is not a constant. It's a curve, and the curve bends down as homes get bigger.
Here's the mechanism. A buyer on a given street is really paying for two different things bundled into one price. They're paying for the location, the schools, the walk to the park, the specific block, which is roughly fixed no matter how big the house is. And they're paying for the physical structure, the sticks and drywall and finishes, which scales with size. The land value and the location premium get spread across every square foot. Put them on a small house and each foot carries a fat slice of that premium. Put them on a huge house and the same premium gets diluted across far more footage, so each foot carries less.
There's a second force pushing the same direction. Every micro-market has a saturation point, a size beyond which additional square footage stops commanding the street's full rate and starts trading closer to what it cost to build. Think about who's shopping a given block. The pool of buyers who want a 3,000-foot house on that street is deep. The pool who want, and can afford, a 6,000-foot house on that same street is thin. When you build past what the block's buyer pool is hunting for, the market doesn't reward those extra feet at the headline rate. It pays you something closer to construction cost, sometimes less, because you've built a house that has to wait for a rarer buyer.
So the largest home in a comp set showing the lowest price per foot isn't an outlier to be scrubbed. It's the curve doing exactly what the curve does. And the small, sharp house on the good street showing the highest number isn't lucky. It's the location premium, undiluted, showing up in full.
Let me make this concrete with illustrative numbers so you can see the arithmetic. Say a street supports a proven finished value of around 400 dollars per foot at a strong renovation tier. A 2,500-foot house at that rate exits at roughly 1,000,000 dollars. Clean and simple.
Now say you own the lot next door and the zoning lets you build 4,000 feet. The spreadsheet-first move is to multiply: 4,000 times 400 equals 1,600,000 dollars, and you go tell your lender the bigger house is worth six hundred grand more. But the street doesn't pay 400 across the board at that size. Past its saturation point, those extra 1,500 feet don't earn the location rate. Say the marginal feet trade at 250 dollars, closer to what they cost to build. Now the math is 2,500 feet at 400 plus 1,500 feet at 250, which is 1,000,000 plus 375,000, or 1,375,000 dollars total. On paper the blended rate just fell from 400 to about 344 a foot.
You spent real money framing, roofing, wiring, and finishing 1,500 extra feet, and the market paid you roughly your cost to produce them and not a dollar of premium. Your gross went up. Your margin got worse, and your risk got worse with it, because now you're holding a bigger, more expensive, more unusual house waiting on a thinner buyer pool. That is how a bigger build makes less money. I've watched investors do exactly this and call it ambition.
Build to the street's proven ceiling, not to the maximum the lot allows.
This isn't theory for me. Look at our deal at 3215 West 83rd Street in Leawood. There's a same-street comp 196 feet away that traded at 1.75 million dollars, roughly 381.85 dollars per foot. One hundred and ninety-six feet. You can throw a baseball that far. That single sale answers the only valuation question that actually matters on this deal: what does this exact location, at this quality tier, pay per foot? Everything else in the comp set is an approximation of what that one sale states outright.
And in that same set, the biggest home posted the lowest price per foot. Right on schedule. If I had ranked the comps by price per foot and reached for the highest number to justify a bigger, richer build, I'd have been reading the curve upside down, pricing in a premium the street pays on smaller homes and assuming it holds as the house grows. It doesn't hold. The 196-foot comp isn't just the closest sale by distance. It's the closest by relevance, because it isolates location and tells me the rate before I've spent a dollar on drywall.
That's the whole reason same-street evidence beats a dozen zip-code averages. A zip-code average blends streets that aren't comparable, sizes that aren't comparable, and finish levels that aren't comparable, then hands you a single number with false confidence. A comp 196 feet away controls for all of it at once. Same schools, same commute, same block, same buyer pool. The only variables left are size and finish, and those you can actually adjust for.
Here's the practical instruction I run every renovation against: match your finished size to the street's proven ceiling, not to the maximum the lot allows. The strongest exit multiple almost always belongs to the home that is fully realized for its exact location, not the one that out-builds its block.
Fully realized is the key phrase. It means you deliver everything the street rewards, the right finish level, the right layout, the beds and baths that block's buyers expect, and then you stop. You don't add the fourth-car garage the block won't pay for. You don't blow out the primary suite to a size the comps don't support. You spend money only where the street has already shown it pays that money back, with a premium. Every dollar past that line is a dollar you're spending at construction cost hoping the market gives you a rate it has never given anyone on that block.
I learned to think this way long before I owned a house. As a kid I sold sweet corn on the roadside with my dad, and we threw an extra ear into every dozen. That extra ear cost us almost nothing and it's the thing people remembered, so they came back and they told their neighbors. But we didn't hand out three extra ears. There's a point where generosity stops buying loyalty and just becomes margin you set on fire. Renovating past the street's ceiling is the three-extra-ears mistake in construction form. You feel generous. The market doesn't pay you for it.
Once you accept that price per foot is a curve, you can use it as a real underwriting instrument instead of a bumper sticker. My background is in credit, I built underwriting at a lender where we pulled all three bureaus and priced risk on hard data, and I bring that same posture to a house. I want to know the shape of the curve on the specific street before I commit, not the average of a hundred streets I'll never touch.
Practically, that means before I buy I plot the street's own sales by size against price per foot and look for where the number starts falling off. That falloff point is the saturation size, and it tells me the maximum footage the location will still reward at a premium rate. I underwrite my finished size at or just under that point. If the lot lets me build past it, I treat those extra feet as trading at marginal cost, not at the street rate, and if the deal only works when I pretend those feet earn full price, the deal doesn't work. I pass. I would rather be paid to be patient than forced to be lucky.
The same lens explains why I'll pay up for the cheapest house on an expensive street and won't touch the nicest house on a cheap one. On the expensive street, the location premium is fixed and high, and I get to buy it at a discount because the house is currently dated. My renovation dollars convert into that premium at full rate. On the cheap street, there's no premium to capture. I could deliver a flawless renovation and the ceiling above me is low, so my best-case exit is a mediocre one. The house isn't the asset. The street is the asset. The house is just my access point to it.
I want to be fair to size, because there's a version of the big build that works and it's worth naming so you can tell the two apart. Size wins when the location itself has effectively no ceiling, when you're on land so rare that the buyer pool for enormous homes is genuinely deep. That's a different game with different rules, and it can be a very good one.
We ran that play once, a ground-up luxury rebuild in Loch Lloyd. Bought the position for 410,000 dollars, put roughly 3.1 million into the build, and sold around 4.23 million for about 720,000 in profit. That worked because the location supported a home at that scale. The buyer pool for a house like that in that spot was real, not imagined. We weren't fighting the curve, we were building on a street where the curve stayed high far longer, because scarcity of comparable land held the premium up.
The mistake isn't ever building big. The mistake is building big on a street that has already told you, in its own sales, that it stops paying the premium at 3,500 feet. The street files that report every time a house trades. Most investors don't read it. They read the zoning, see the maximum square footage the lot allows, and mistake permission for demand. Permission is what the county grants you. Demand is what the block's last few buyers actually paid for. Only one of them shows up at your closing.
Strip it down to a workflow. Before you commit to a renovation or a build, pull the same-street sales first and weight them above everything else, with the nearest and most recent sale carrying the most weight. Plot those sales by size against price per foot and find where the rate starts to fall, and treat that falloff as your finished-size ceiling. Underwrite your project at or below it. Any footage past that point, price at marginal construction cost, not at the street's headline rate, and if the deal only pencils when you pretend the extra feet earn full price, walk.
When the largest comp in your set posts the lowest price per foot, don't delete it. Circle it. It's the market drawing the exact curve you're about to build against. The small house with the fat per-foot number is telling you what the location is worth undiluted, and the big house with the thin one is telling you where the location stops paying for size. Read both together and you know, before you spend a dollar, exactly how much house this street will actually reward. That reading is worth more than any zip-code average will ever give you, and on the good streets, it's the whole edge.
Strategy notes on underwriting, structure, and disciplined execution. No noise.