A bid is a price attached to a scope, and the scope is where the money hides. The $38,000 framer left out the stairs, assumed you supply the hardware, and priced a two-week window that will slip. The $44,000 framer included all of it and pads for a schedule they can actually hold. Compare the two numbers and you learn nothing; compare the two scopes and the “expensive” bid is often the cheaper job. The cheapest bid is simply the one that forgot the most — and you inherit everything it forgot at change-order prices, which are the worst prices you will ever pay.
Leveling means normalizing each bid to one common scope so the only thing that varies is the price. Build a single line-item list of what the job actually includes — every task, every material responsibility, every allowance — and map each bid onto it. Where a bidder excluded a line, add your best number for it back into their total. Now the totals are honest: you are comparing the same house, not three different guesses about what “framing” means.
Do this and the spread usually collapses. The $38k and $44k bids, once both include the stairs, the hardware, and a real schedule, land within a few thousand dollars — and the decision moves from price to the thing that actually matters: who you trust to show up and hold it.
The line that costs you is the one that isn’t there. Before you compare totals, list every bidder’s exclusions and assumptions side by side: who supplies what, what’s allowance versus fixed, what schedule is priced, what happens on a change. A bid with no exclusions section is not a clean bid — it is an unstated one, and the omissions surface on site. Ask for the exclusions in writing; a sub who can’t tell you what they left out hasn’t finished bidding.
Two subs level to the same number. Which do you award? This is where your own record earns its keep. The sub who came in on his last two jobs, answered the phone when the inspection moved, and left the site clean is worth a premium over a stranger with a slightly lower bid and no track record. Keep a rating on every vendor you’ve worked with — how the work actually landed, not how the bid read — and the award decision stops being a coin flip. A rating earned on your jobs can’t be gamed the way a reference can.