Bid analysis: what you charge, what it costs, and what wins

Pursuit → Bid Analysis is a new surface built from the bids you've already priced:

• The bid register — every project's latest estimate on one row: bid date, GC, status, won/lost, bid price, direct cost, and gross margin (graded with the same 15% / 8% thresholds as the estimate builder). Filter by outcome, peek a row for the labor/material/equipment/sub split, and export the lot to CSV. A small Δ flags any project whose recorded bid value has drifted from what the estimate actually totals.
• Overall performance — bids, decided, win rate, bid volume, and won-vs-lost margin, value-weighted so a big job counts like a big job.
• System benchmarks — for every assembly you've bid from the library: how many bids it appeared in, how many won, and its weighted-average price and cost per unit, with the min–max spread. The same rollup runs per cost code × type × UOM, so lines outside assemblies are benchmarked too.
• Per-bid breakdown — open any bid to see its unit prices next to your org benchmarks, line group by line group and assembly by assembly, with the delta called out.

Benchmarks pool approved and submitted bids plus decided pursuits, base bid only — drafts on live pursuits don't muddy the averages, and TBD-tagged lines stay out of the unit math. Every number comes from the same pricing formula as the estimate builder footer, so the two can't disagree.