How much do contractor bids actually vary? We pulled 20 real public bid records to find out

"Get three quotes, because they'll vary a lot" is standard advice for hiring a contractor. Nobody attaches a real number to it. Public construction bids are one of the few places this question has an actual, disclosed answer: government agencies in the US are generally required to run competitive bidding on construction work and to publish the results, bidder by bidder, in dollars. So instead of repeating the folk wisdom, we pulled 20 real, awarded public bid tabulations ourselves and measured the spread directly.

Every project below is a real, completed public solicitation with a published bid tabulation naming every bidder and their amount. Sources: Louisiana's Division of Administration, Facility Planning & Control; the University of Kentucky's Purchasing Division; Fairfax County, Virginia and Fairfax County Public Schools; and New York City's Department of Citywide Administrative Services. Several of these agencies' own sites currently block automated access; where that happened, we confirmed the identical document through the Internet Archive's Wayback Machine rather than estimating from a secondary source, and we note it below.

In every one of these five, the lowest bidder won. The Claiborne building drew the most competition of any project in our sample, nine bidders, and also the widest roofing spread: the highest bid ran 86% above the lowest.

The Fairfax County line is a different kind of contract than the other five: an hourly-rate, on-call term contract rather than a single fixed-price job, so its 3.41x reflects very different vendor labor-rate structures rather than one project's true cost uncertainty. We include it because it is real and disclosed, but weight it separately from the fixed-price projects above it.

The widest spread in the whole dataset outside the term contract above is the New York City electrical citywide contract, where the higher of two bidders came in more than double the lower one. With only two bidders, a single outlier estimate has nowhere to average out, which is likely part of the story.

Across all 20 verified projects, ratios ranged from 1.02x to 3.41x, with a typical spread around 1.4x, meaning the highest bid usually ran about 40% above the lowest. And in every single verified case, the agency awarded the contract to the lowest bidder. We found zero examples in this sample of a 'best value, not lowest cost' award overriding the cheapest number.

For public hard-bid work specifically: a 25 to 65% spread between qualified bidders on the same job is normal, not a sign that someone made a mistake, and the agencies in this sample awarded strictly on price every time. If you bid public contracts and consistently land near the middle of that range, being the low bidder more often is worth pursuing directly, since that is functionally the whole selection criterion.

For a homeowner or business owner privately collecting quotes, the process is different enough that these exact ratios should not be quoted as your expected spread. What does carry over is the underlying shape: real, qualified contractors bidding the same defined scope of work still land tens of percent apart from each other. A wide gap between two quotes is not automatically evidence that one of them is wrong.