Ask five techs how they size a system for a house they haven’t measured yet, and at least a couple will give you some version of “500 to 600 square feet a ton.” It’s a fast number, and it’s wrong often enough to matter. It doesn’t know the house has single-pane windows facing west, a ceiling that goes to 14 feet in the great room, or that it sits in a climate zone where the design temperature swings 20 degrees from the neighborhood three towns over. Square footage alone can’t carry that information, and a system sized on square footage alone is a guess wearing an invoice.
The name for the thing that actually accounts for all of it is Manual J, the ANSI-recognized residential load calculation standard published by ACCA, the Air Conditioning Contractors of America. It’s not a competitor’s marketing term. It’s the industry’s own standard, and in most jurisdictions that have adopted the IRC or IECC as their building code, a Manual J calculation, or an approved equivalent, is what the code actually requires before a system gets sized for a permit.
What a skipped Manual J actually costs
Oversizing shows up as comfort complaints, not obviously as sizing. A too-big system blasts a room to setpoint fast, shuts off, and starts the cycle again a few minutes later. That short-cycling pattern never runs long enough to pull real humidity out of the air, so the house hits 72 on the thermostat and still feels damp and clammy. The customer doesn’t call and say “my system’s oversized.” They call and say the new unit doesn’t work right, and that call lands on the shop that installed it, not the shop that guessed.
Undersizing shows up on the worst possible day. A system sized a half-ton light keeps up fine on a mild afternoon and falls behind the instant a real heat wave or cold snap hits, which is exactly when a callback is hardest to staff and worst for the review it generates.
Both directions waste money that’s invisible until it isn’t. An oversized system also runs less efficiently over its life, and a system sized off a guess instead of a load calc gives a shop no paper trail if a customer, an inspector, or a warranty claim ever asks how the tonnage was decided.
Why this is a business question, not just a technical one
Any shop that runs a real Manual J on every estimate is doing something a lot of competitors skip, and it’s the kind of thing a homeowner can actually see. A printed load calculation attached to a quote reads differently than a number scribbled on the back of a business card, especially to the customer who’s already gotten two other estimates that were both just “a 3-ton unit should do it.”
It also means fewer comfort callbacks per install, which is the quiet cost center that never shows up as a single bad number, just a slow bleed of truck rolls, warranty labor, and the review that mentions “never felt right after they installed it.” A shop that can point to the calculation when a customer questions the sizing has an answer. A shop that guessed has an apology.
Where to actually get this right
ACCA publishes the Manual J standard itself, along with Manual S (equipment selection) and Manual D (duct design), the three that are meant to be used together rather than Manual J in isolation. Their technical manuals page is the primary source, and it’s worth having a tech actually read the standard rather than relying on whatever a supply house rep summarized secondhand. Most major load-calculation software (Wrightsoft, Cool Calc, Elite) implements the Manual J procedure directly, so the barrier for a shop that isn’t doing this today usually isn’t the standard, it’s just making the calculation part of the estimate instead of an afterthought.
The house doesn’t care what the rule of thumb says. It cares whether the equipment matches its actual load on the hottest and coldest days of the year. That’s the whole job. The square-footage number was never going to answer it.
