A small firm cannot read a dozen bid portals by hand, every morning
Public agencies do not post owner's-representative and construction-management work in one place. It is spread across a state's own procurement site, a county's purchasing portal, and a handful of third-party bid-notice services, several of which sit behind a Cloudflare challenge or a login wall to anything that is not a signed-in browser session. A person can read all of that by hand, once. Doing it every weekday, across every relevant state and county, is not a job to give a person twice.
We built a repeatable sweep instead: a run against the public search endpoints that will answer a plain keyword query without a login, a logged-in pass against the portals that require one, and a standing note of which sources are walled off entirely so nobody re-discovers the same dead end twice. Every hit gets scored GO or NO-GO against one test: does the scope match a licensed, advisory owner's-representative or construction-management-services positioning, and is there enough runway left to actually build a response. The reasoning for every call, GO or NO-GO, gets written down next to the listing, not just the verdict, because the reasoning is what makes the next sweep faster than the last one.
The scope language that read like our own pitch, and the fact that mattered more
A public affordable-housing RFP for construction-manager and owner's-representative services on a 61-unit development, in the $65 to $75 million range, came through the sweep in mid-2026. Its own definition of the role was unusually narrow: an advisor contractually obligated to the housing partnership's interests, who does not direct the general contractor, subcontractors or designers. That is close enough to an existing, licensed advisory positioning that it read as a rare fit, and a full nine-section proposal, an itemized fee translation and a reusable review-artifact demo were all built inside two days, comfortably inside the deadline.
What could not be produced by working faster was a fact: the RFP required three named, recent, contactable client references at the exact scope of the work, worth twenty of the rubric's hundred points, and that many qualifying references did not yet exist. Three more days passed after the build finished before the decision to pass was made, and the honest accounting is not that time ran out. It is that no number of additional days manufactures a specific, recent client relationship that is not already in motion. The reference requirement is a supply question, not a writing question, and it was checked fifth, after four sections had already absorbed real construction effort, instead of first.
How we came at this one
The question asked, once the gap was confirmed, was not whether the proposal was good. It was the score arithmetic: a zero on a twenty-point criterion caps a submission at eighty of a hundred, while an ordinary competitor who clears that criterion and scores an unremarkable eighty percent on the rest lands at eighty-four, above the ceiling this submission could reach. That comparison assumes a competent, responsive field exists, which was never actually checked, so the pass decision accepted that uncertainty rather than resolving it, and weighed it against a second cost: a documented, weak submission to a specific agency costs more against a future pursuit with that same agency than not bidding this one does.