Kanopi Bid

A bid in your format. A copy your client can read.

Kanopi prices the takeoff the way your shop prices, with your subs, your rates and your markup, and hands back a finished bid in your own template.

For the contractor sitting at a homeowner's table: every line says where its number came from, and the owner copy is written in plain words, so the conversation is about choices, not acronyms.

Kanopi Bid in about 90 seconds. No sound needed: the captions are on screen.

How we do it

How a Kanopi bid is built.

  1. 01

    Learn how your shop prices

    You send two or three bids you won. Kanopi reproduces one of them without seeing it, and we compare line by line. No live bid ships until that passes.

    For the estimator

    Calibration reconciles on your rates, your markup stack and your scope boundaries until the total lands inside the band for its plan stage. A total that lands by compensating errors does not pass. Thin building types are pulled toward the wider history rather than trusted on one job.

  2. 02

    Price from sources you can see

    Every rate carries where it came from: a sub's quote, one of your won bids, a vendor price book or a published reference. A rate only counts as verified with its source and date beside it.

    For the estimator

    One versioned rate library. Each rate has a unit, a basis, a range, a confidence level and a last-verified date; local sales tax and permit tables sit beside the rates. Your history and your subs come first; the library fills what they do not cover.

  3. 03

    Check it against a partner's book

    Where a trade partner or supplier shares their price book, every line is compared with theirs and marked as a match, worth a look, or a conflict.

    For the estimator

    Partner overlays return Kanopi's number and theirs side by side with a MATCH, REVIEW or CONFLICT verdict; within 10 percent is a match.

  4. 04

    Gates that stop a bad number

    The engine refuses to finish a bid with a quantity it could not resolve or a rate it does not know. Those go to a person, not into the total.

    For the estimator

    Hard gates block on unresolved areas, failed invariants and unknown rate keys; softer gates send disagreements, out-of-band lines and building types not yet calibrated to review.

  5. 05

    Tag every line

    Each line is marked Quoted, Takeoff, Assumed or Allowance, so a reviewer can go straight to the assumptions and spend the review there.

    For the estimator

    Lines are organized by CSI MasterFormat division. Quantity confidence and rate confidence are kept separately, and the line shows the weaker of the two.

  6. 06

    Two copies, one estimate

    You get a working copy with the basis behind every line, and an owner copy in plain words that your client can follow. Same numbers, two readers.

    For the estimator

    The owner copy leaves off the internal basis and the margin-of-error bands, which stay on the working copy, set by plan stage on labor and self-performed work.

  7. 07

    Every bidder prices the same job

    When you are the one collecting bids, we write one baseline scope from our takeoff and send it to every bidder, so the bids come back comparable.

    For the estimator

    Each takeoff line goes out with an include checkbox and a comment field; adds and deducts are itemized apart from the base. Leveling then compares like with like.

The tools

Where the numbers come from.

A bid is only as good as its sources. These are the ones Kanopi prices from, in the order it trusts them.

Your won bids
How you price: your markup stack, your scope boundaries, your conventions.
Your subs' quotes
Tagged Quoted, carried as given. In our blind study, vendor quotes were reproduced exactly.
Partner price books
Loaded as overlays and compared line by line.
Sourced rate library
Vendor and market references with links, each with a last-verified date.
Local tables
Sales tax and permit fees by city, beside the rates.
The Kanopi takeoff
Measured, overlaid and blind-checked quantities, so every line is quantity times rate.

How accurate

How close, said honestly.

The counting is solid. The price has to learn your shop first. Here is how far off it was, measured the hard way.

  • Held-out accuracy: 18.6 percent mean error and 31.9 percent worst, on single-family bids, as of June 10, 2026.
  • Before calibration, the 11-bid blind study totals landed a median 35 percent from the shipped bid. Four of eleven fell inside the band for their plan stage.
  • Every miss outside the band traced to a pricing convention (scope tier, who supplies material, the furniture boundary, the markup stack), never to a measurement.

How measured: held-out means leave one out. We calibrate on every bid but one, price the one left out blind, compare, and repeat. In-sample matches are never quoted as accuracy. The blind study used a standard rate library with no tuning to the shop.

Next to other tools

Kanopi Bid is how the price gets built and checked before the job starts. Other tools do other parts of the job well.

RSMeans and other cost books
National unit costs with city factors, the same for every shop. Kanopi prices from your history, your subs and sourced rates. A cost-book cross-check by division is on our roadmap, not built.
Estimating software
Gives your estimator a cost database and a template to build the bid in. Kanopi builds the bid and shows the basis for every line.
Procore
Ahead of Kanopi after award: budgets, commitments, change events, forecasts and invoices. Kanopi has none of that. It hands the number to the system you already run.
Leveling by spreadsheet
Most bid leveling happens after the bids arrive. Kanopi writes the baseline scope first so they arrive comparable. A push-button leveling sheet is on our roadmap, not built.

Vendor descriptions are taken from each vendor's own pages, checked October 2026. No competitor prices are shown.

Send one project, a bid you already sent, and how you price. The first bid is free.

Try your first bid free