Use this article when Estimate Setup is complete for a product category and you need to prove the category quotes correctly across the quantity bands the shop sells. After this you will have a repeatable matrix of test estimates that exercises routing, production cost, and commercial price for every revenue-critical category.
Before you begin
Every setup edit you intend to test is already Applied through Pending Changes. The Pending Changes count in the sidebar reads zero.
You have a reference price for the same specs — usually an MIS export, a current spreadsheet, or a printed sample quote. This can be approximate; exact parity is not the goal.
The finance lead and the implementation lead have agreed an acceptable percentage band per category (typical band: ±5%).
1. Build the comparison matrix
The matrix is the artifact you maintain across the validation run. Build it in a spreadsheet or shared doc with one row per spec × quantity. Fill in the fields below.
Per category (one row per spec × quantity)
Category — the product category being tested, e.g., Stitched Book, Folded Leaflets, Banner.
Finished spec — the size, page count, colour, substrate, and finishing options that match how the shop quotes the category. Use a realistic spec, not an edge case.
Quantity — the test point. Cover the bands the shop actually sells. A typical pattern is
500 / 1,000 / 2,500 / 5,000, with one larger band (e.g., 25,000) for run-length-sensitive categories.Reference price — the price the shop's MIS or spreadsheet would produce for the same spec at the same quantity.
Estimator price — left blank now; filled after running the quote.
Step notes — left blank now; recorded after reading the price breakdown.
Pass / fail — left blank now; marked after comparing inside the agreed band.
Cover the top revenue categories first. Do not try to cover every SKU.
2. Run a baseline quote per matrix row
Open Manage Estimates in the sidebar, then choose Create new estimate. Fill in the fields below.
Quote creation
Customer — the customer the shop will use for production. A dedicated test customer is fine if the team prefers to keep validation quotes off the live customer's history.
Product category — the category being tested.
Spec — the finished spec from the matrix row: size, page count, colour, substrate row, finishing options.
Quantity — the matrix-row quantity.
Wait for Estimator to recalculate. Open See price breakdown and record on the matrix row: the route (which press), the substrate cost, the make-ready and run times on Print, Cut, Fold, and Bind steps, and the bucket totals. Repeat for every quantity column on the row.
The Final price updates on the quote totals each time you change the quantity.
3. Isolate production cost (optional, recommended on first pass)
On the first pass through a category, temporarily set every bucket markup percentage on the category to zero. This makes the breakdown show production cost without margin policy on top, so a discrepancy can be traced to a step (machine speed, substrate price, conditional-step assignment) without the markups confusing the picture.
Open Estimate Setup → Products → Categories, find the category row, set every bucket markup to 0%, Save, and Apply Changes. Re-run the matrix quotes and record the production-cost values. Restore the markup percentages before category sign-off — validation is not finished until the commercial price is checked.
The Pending Changes count increments by one when you set markups to zero and returns to zero on Apply Changes.
4. Record results
For every matrix row, compare Estimator price to reference price. Mark the row as pass when the result is inside the agreed band. Mark as fail when it is outside, and record the step name on the row (Print, Cut, Fold, Bind, Substrate, or a custom step name) where the divergence first showed in the breakdown. The step name points at which setup layer to debug — production setup for steps, pricing setup for bucket markups or rules. See Why a pricing rule did not apply when the divergence sits at the markup or rule layer.
5. Sign off the category
When every row for a category passes inside the agreed band, initial the matrix and move to the next category. Use the Setup validation checklist for the per-category sign-off form.
The Final price on a re-quoted matrix row now matches the validated number, and the row carries a pass mark in the comparison matrix.
Things to know
Multi-product tenders are validated through the multi-product quote flow, not by stacking single-product validation quotes. The single-product matrix described here is the right starting point for each category individually before the team moves to combined-product tests.
The AI Configurator is useful for spot-checking a spec after the baseline matrix is green — paste a spec and confirm the chat-driven quote matches the manual quote within the same band. It is not a substitute for the structured matrix.
A failed row with markups zeroed but production cost still wrong tells you the issue is in production setup (machine speeds, substrates, finishing). A failed row that passes when markups are zeroed but fails when markups are restored tells you the issue is in pricing setup (markup percentages or a pricing rule).