Every government auction is a snapshot of the broader market at a given moment in time. We believe these auctions exhibit structure, that the competition within them leaves patterns, and that pricing behavior is measurable. Mostly, we believe that the underlying market structures can be discovered and learned.
“Researching competitive behavior in construction auctions and developing tools that help heavy civil contractors better understand their markets.”
We forecast the projected winning bid before the letting.
Models are trained by state and work category, and every forecast is locked before bids open and scored against the published result. 409 settled contracts across Florida, Kansas, and Texas. The record is public, including the misses.
ForecastingYears of bid tabs, parsed. Who bid what, in which district, on which job type. Full item-level history for every contractor in your state.
The Research Center allows for quick contract, vendor, district, and item-level pricing lookups.
Market AnalysisThe model prices the market. It does not know your production rates, your backlog, your crews, or which job you want badly enough to sharpen your pencil on.
Put your team’s number next to the market’s and see the gap before bid day, not after.
Bid DeskContracts load with current quantities and a forecast, before your team opens a plan set. You start the letting knowing roughly where the market is likely to sit on every job.
You have twenty jobs and time to price three properly. Typical spreads and field sizes tell you which ones are worth real effort and which were never winnable. Spend more time on the jobs that matter most.
Put your number against the market’s. See where you sit relative to a risk-neutral bidder, what the runner-up spread has looked like on jobs like this one, and how much room there was the last time this district let this kind of work.
