ARV Analysis, Step by Step — the Discipline Behind the Number

ARV (after-repair value) is what the property should sell for once it's fixed up. Get it right and every downstream number (your max offer, your rehab budget, your profit) has a foundation. Get it wrong and you're doing confident arithmetic on a fantasy. Here is the discipline, start to finish, the way you'd do it by hand.

Step 1: Closed sales only

Not active listings, not pendings, not "contingent." Those are asking prices, and asking is free. A sale that closed and recorded is evidence; everything else is a negotiation in progress. This single rule removes the most common source of inflated ARVs.

Step 2: Match the house, not the neighborhood average

Same property type first (a townhouse is not a comp for a detached ranch), then bedroom count within reason, then size — you want comps in the same general square-footage band as your subject. A much bigger home that sold high mostly sold high because it's bigger. Distance matters too: the tighter the radius, the fewer school-district and block-quality surprises you inherit.

Step 3: Recent, or explain why not

Prefer the last several months. If you have to reach back further to find matches, that's a signal in itself — thin sales activity means every conclusion deserves less confidence, not a wider net and the same confidence.

Step 4: Normalize to price per square foot, then hunt outliers

Divide each sale price by its size and line them up. Now interrogate the ends of the list, not the middle. The $210/sqft sale in a $150 neighborhood: full gut renovation? Corner double lot? A sale between family members at a courtesy price? Every outlier has a story, and the story decides whether it belongs in your average or in the trash.

Step 5: Compare condition to YOUR after-repair state

ARV assumes your subject is renovated. So the comps that matter most are the renovated sales, not the tired ones. If half your comps are dated interiors, your "after-repair" value is quietly a "before-repair" value and your profit math is broken from the start.

Step 6: Average the survivors, sanity-check the result

Average the $/sqft of the comps that survived, multiply by your subject's size, then step back: does the number pass the smell test against the ZIP's general price level? A result wildly above the area's normal ceiling needs an explanation you can say out loud to a lender without wincing.

Step 7: Be willing to say "not enough data"

Under a handful of genuinely comparable closed sales, the honest answer is that the ARV is uncertain: widen carefully, or price the uncertainty into a lower offer. The discipline is admitting it, and most tools (and most sellers' agents) won't.

Step 8: Stress it before you sign anything

Knock a few percent off the ARV and add some to the rehab, and look at the deal again. If it only works in the sunny version, it doesn't work.

Done by hand, this is a solid half hour per address — which is fine for one house and impossible for a whole market. That's the entire reason deal software exists: not to replace this discipline, but to run it in seconds and show the work so you can audit it. That's the standard we hold our own ARV engine to (comps used, comps rejected, reasons attached) and the standard you should hold any tool to, including ours.