All articles
estimating

Bid leveling: making five subcontractor bids actually comparable

You get five window and glazing bids back for the same project. Surface scanning says one is clearly the low number. Then the estimators sit down and realize every one of the five priced a different scope — one carried the storefront, one left it out, one included the balcony railings nobody asked for, and two of them rolled interior-caulk allowances in while a third excluded them.

The raw low number is not the low number. Bid leveling is the step that makes them comparable, and skipping it is how a GC wins a job with a number its lowest sub can’t actually honor.

Stylized 3D of five open bid documents stacked at angles, a neon lime-green alignment line drawing across their scope lines, on a deep navy background

The short version

Leveling means normalizing every bid back to the same scope baseline, then pricing the differences. You decide what the “apples” are — the exact assemblies, quantities, and inclusions your drawings actually call for — and you pull that same baseline through every bid. Anything a sub added or left out is priced and added or subtracted, so the comparison happens on even ground.

You do not take the low number to the award. You take the low leveled number.

Build the baseline before the bids come in

The baseline is the scope agreement, and it has to exist before you open the quotes, or you will just compare five moving targets.

  • Write the scope of work down. Unit by unit: glazing systems, glass types and thicknesses, door hardware, perimeter sealants, anchorage, shop drawings, permitting, delivery, hoisting, and field labor. If it is on the drawings, it belongs in the baseline.
  • Fix the quantities you will level against. Use one takeoff source for all five bids. Do not level against each sub’s own takeoff — that double-counts or drops scope at random. Your schedule-of-values discipline and a consistent window/door schedule takeoff give you the neutral set of numbers to apply to everyone.
  • Pick the alternates up front. Decide which scope items are base and which are alternates before comparing, so a sub who priced the alternate inside base is flagged as an outlier rather than silently more expensive.

Where the bids diverge

Five bids on one scope set rarely disagree on the glass itself. They diverge in five predictable places, and those are exactly where leveling earns its fee.

  1. Scope inclusions and exclusions. One sub carries the storefront as a line item, another excludes it as not-in-scope, a third buries it in a lump sum. Every one has to be re-priced against the baseline.
  2. Different quantities. Subs take off different amounts because they interpret the drawings differently. Level against your numbers, not theirs.
  3. Allowances. Interior caulk, sealants, miscellaneous metal — some roll these in, some list them as allowances, some just leave them out. An allowance line that only exists in half the bids is not a fair basis for comparison.
  4. Alternates inside base. The sub who prices an alternate inside their base looks higher than the one who left it out. Flag the difference as a line, not a verdict.
  5. Terms, not price. Lead time, delivery, hoisting, and whether install labor and supervision are in — the arithmetic of a bid is only half the story. Two bids landing at the same number can still have wildly different work and risk in them.

Price the differences, don’t just note them

The discipline that separates a level from a spreadsheet is that you assign a number to every delta and reflect it in the total, rather than appending a comment like “note: does not include storefront.”

That note is exactly the trap. A “note” does not move the number, so the comparison is still false. To level, you apply a carry cost — the value your team assigns to the omitted or extra scope — to the bid total so every bid ends up on the same basis.

When you cannot price a delta with confidence, do not guess. Mark it as an open item, assign it to the appropriate party to resolve, and keep it visible through the award decision rather than burying it in a leveling tab that nobody opens after the comparison meeting.

The two numbers that matter

After leveling, you hold two numbers for each bid:

  • The bid as quoted — what the sub wrote down, which is what you would owe them for their stated scope.
  • The leveled total — bid-as-quoted adjusted by your carries to the common baseline, which is what the bid is worth on the project.

The second number drives the comparison. The first only tells you whether to double-check the arithmetic. Comparing leveled totals across the five bids surfaces the genuinely competitive subcontractor, and it gives you an honest number to defend in your overall bid when the owner or the estimator above you asks how you reached it.

Why a sub’s number should never move your scope

The natural temptation is the reverse: let the low bid define the project. If the low number excluded the storefront, the instinct is “great, that’s the quote — let’s just run with it.” That is how you win a job with a scope you have not fully priced, and it comes due at the worst possible moment, usually as a change order the day before glass is scheduled.

Leveling keeps the scope as the constant and the bids as the variable. The scope is decided by the drawings and the contract, not by whichever sub happened to leave something out.

Make it a repeatable step, not a meeting-day scramble

Leveling is not glamorous and it does not need to be dramatic. It needs to be a fixed step in the estimating process with a defined output — a leveling sheet where every bid has a total, every delta has a carry, and every open item has an owner. If it only happens in a three-hour meeting the morning bids are due, it will get skipped the moment the schedule tightens, because everyone is busy reconciling the very exclusions the sheet was meant to catch.

The estimating team that builds this once and reuses the baseline across bids gets two things: faster comparisons and fewer surprises. A structured estimating workflow and a pricing model that assumes leveled, comparable subcontractor pricing are what let a GC sit on a real low number instead of a false one.

Level the bids in the office. The field will thank you when the storefront shows up in someone’s scope and it is covered.


Ready to talk through how a leveled subcontractor comparison should fold into your estimating process? Get in touch.