Definition
A bid outlier is a number that doesn't belong with the rest of the field — either a whole subcontractor bid whose total sits well above or below the others, or a single line item priced far from what every other bidder quoted for the same work. When a general contractor opens 5 to 15 bids on a scope package, most cluster within a believable band of one another. The one that breaks away is the outlier, and it is usually a symptom of something specific: a misread quantity, a decimal-point typo, a unit mix-up, or scope a bidder simply left out of the price.
An outlier is not automatically a bad bid, and a low outlier is not automatically a bargain. It is a flag that says this number deserves a second look before you trust it. The cheap surprise that wins the job on the cover sheet is the same number that comes back as a change order once the missing work surfaces. Spotting outliers early is how a GC avoids awarding to a price that was never real.
A worked example
You receive eight bids on a drainage package. Seven land in a tight band around $1,000,000 — $960,000, $985,000, $1,010,000, $1,020,000, and so on. The eighth comes in at $750,000, a full 25% under the field. Because that is 20% or more below the median of the eight totals, it flags as a total-bid outlier. You read the low bid line by line and find the reason: the bidder carried no structural excavation and backfill, roughly $260,000 of work your SOW requires. Add it back and the scope-equalized total is about $1,010,000 — right in the middle of the pack. The $750,000 was never the real price; it was an exclusion wearing a discount.
Why it matters when you evaluate sub bids
Outliers are where bad awards begin. Rank the raw totals and the lowest number wins by default, which means the bid with the biggest hidden gap looks the most attractive. Catching outliers is what separates a defensible award from a coin flip: a flagged total tells you which bid to scrutinize before it reaches the recommendation, and a flagged line item tells you exactly where the price stops being trustworthy. The same discipline protects you on the high side too — a line priced at twice the field median can quietly inflate a total or hint that a bidder misread the plans. Whether you are leveling the field, scoring it, or writing the award memo, you cannot stand behind a number you haven't tested against its peers.
How Bid Reasoner handles it
Bid Reasoner runs four deterministic pricing rules across every bid that has been read, two of them built to catch outliers directly: a total-bid outlier 20% or more from the median total (3+ bids), and a peer outlier where a line item runs 2x or more, or 0.5x or less, of that pay item's median (3+ bids). Those rules fire the same way in any US state, because the comparison is anchored to peer-median normalization rather than government data. The other two rules — unbalanced unit prices at or under $1.00 and front-loaded mobilization at 10% or more of total — catch the distortions that hide inside a total that looks normal. Every pricing-rule flag shows the numbers that tripped it (an outlier flag shows the price or total against the median it was measured against), and the two-way scope-coverage gap analysis lists the missing structural excavation (by pay-item code) behind a low outlier so you can see why the number is off. The flag adds to the bidder's risk score on the Compare page, and if you award against the recommendation, the award page asks for a category and a reason, and both print in the award memo.