Definition
Bid leveling is the process a general contractor runs on the subcontractor bids it receives for a scope package, restating each one against a single scope of work so the numbers are actually comparable. Subs quote in different formats — one sends a lump sum, another itemizes 40 unit-price lines, a third bundles work under different names — so the cover-sheet totals rarely measure the same thing. Leveling reconciles those differences: it maps every line item to your SOW, adds the cost of anything a sub excluded, and converts units so a per-cubic-yard price and a lump sum for the same work can sit side by side. The result is a set of effective, scope-equalized totals you can rank.
A worked example
You receive two bids on a foundation package. Sub A quotes a lump sum of $640,000. Sub B sends unit prices totaling $612,000 — the apparent low bid by $28,000. But reading Sub B's bid line by line, you find it carries no dewatering, which your SOW requires. Add the missing dewatering at the field rate of about $35,000, and Sub B's scope-equalized total becomes $647,000. After leveling, Sub A is the cheaper bid by $7,000, not the more expensive one by $28,000. The raw totals had it backwards.
Why it matters when you evaluate sub bids
On a heavy-civil package with 5 to 15 bidders, the gap that costs you is the line a sub quietly left out. Comparing raw totals rewards whoever excluded the most, because every exclusion is a future change order with that sub's name on it. Leveling is what turns "lowest number on the cover sheet" into "lowest responsible price for the full scope" — and it is the foundation every later step rests on. You cannot honestly score price, run a scope-gap analysis, or defend an award to an owner until the bids are on the same scope footing. Leveling is the reconciliation; the bid tab is just the grid that displays the leveled result afterward.
How Bid Reasoner handles it
Bid Reasoner normalizes the line items on every bid to your scope of work. It reads each bid PDF, maps the items to your SOW by pay-item code — with a peer-median baseline taken from the bids themselves, so no government data is required in any US state — and lists missing items, like Sub B's missing dewatering, as scope-coverage gaps; scope findings the AI saves cite the document and page. The four deterministic pricing rules then fire on every bid that has been read: unbalanced unit prices at or under $1.00, peer outliers at 2x or more or 0.5x or less of the pay item's median (3+ bids), total-bid outliers 20% or more off the median total (3+ bids), and front-loaded mobilization at 10% or more of total. The result is a leveled tab you can rank under the decision mode you pick.