Every subcontractor Bailey-Harris has rated on a closed job, with the full rating history behind each average. Use it during pricing and buyout: filter to the trade you are buying, then read what the teams actually wrote.
One rating per subcontractor per job, entered at closeout or whenever the crew comes off the job. What you write in the notes is the part buyout actually reads — be specific about people and causes, not just the score.
Entered here recently
Rate a whole job in one pass. Every column here asks the same question Add a rating · single asks, one row per subcontractor — so a bulk import lands exactly the same record as typing them in one at a time. Paste from Excel or pick a file; the headings on your existing Sub Ratings workbooks are recognised too. Nothing is saved until you have read the preview.
The prequalification file behind each subcontractor — safety record, bonding and financial standing. These fields mirror Bailey-Harris's Subcontractor Qualification Statement so what you enter here matches what the sub submits.
Prequal on file
Certified subcontractors available for participation goals. A classification only counts on a bid when the certificate is current, so the expiry column is the one to read first.
Corrections to ratings already on file — a CSI division coded to the wrong trade, a scope description that says nothing, a contract amount that never got filled in. The imported spreadsheets are left untouched; what you change here is stored as a correction on top of them, so it can always be undone and you can see who changed what.
One subcontractor, several spellings
The same firm gets typed a dozen ways over ten years — Jack Ellison, Jack Ellison Paint, Paint / Jack Ellison. Left alone, one sub's record is split across six, and none of them shows the full picture. Merging them keeps every rating and every old spelling; it only decides which name they hang under.
Nothing here is guessed at. The merges already applied were misspellings, punctuation, or the scope / subcontractor format from bid-day sheets. Anything where two firms might genuinely share a word is below, unmerged, waiting for someone who knows the vendors.
Waiting on you
One pair at a time. If they are the same firm, set the name you want them to keep and merge; if they are not, mark them different and the pair will not come back. Pairs that share only a single word are hidden by default because most of them are coincidences — Ellison Painting and Ellison Plumbing & Piping are two different companies.
Names ComputerEase cut off
ComputerEase stores a vendor name in 30 characters and drops the rest — which is why Alexander Electric Company, In and Brockway Building Services, LL read the way they do. The full name is not in the ERP to recover, so these have to be typed once. Every sub whose name is exactly 30 characters is listed; some genuinely are that long, so mark those as fine and they will drop off.
Already merged
Rename any of these if the wrong spelling ended up as the canonical one, or split a spelling back out if it does not belong.
The rating scale
What each score means
| 5 | Best in class. Beat the schedule or the scope, ran themselves, made the job easier. First call next time. |
| 4 | Above expectations. In and out, no problems, minor issues handled without escalation. |
| 3 | Met expectations. Did what the subcontract said. Normal amount of pushing. |
| 2 | Below expectations. Manpower, coordination or quality problems that cost BH time or money. |
| 1 | Do not use. Failed to perform, unsafe, or the relationship broke down. |
How to read an average
An average across two jobs is a hint; an average across eight is a track record. The rail shows the job count next to every score for exactly that reason — check it before you act on a number.
A single 1 or 2 buried under good scores is worth opening. That is usually where a named person, a schedule failure or a safety problem is written down, and it will not show in the average.
Twenty-six company names carried a *** prefix in the source sheet. Nobody knows what it recorded and the data ruled out every obvious reading, so it is ignored here — it affects no score, filter or sort. The prefix itself is still in the stored rows if the meaning ever surfaces.
Where this data came from
What is missing
Coverage by division
How much rating history sits behind each scope. A division with a handful of entries is not proof we have no history there — it usually means the work got coded somewhere else.