Traceable inputs
Price date, tax basis and currency stated.
Bulk, packaging, batch size and loss — costed on one consistent basis.
Drag the example on the right to see how cost and price respond.
01 / From cost detail to client quote
Each packaging option is costed separately, and quotes cite a confirmed cost snapshot. Click to view product screens.
Product screen: cost breakdown by bulk, packaging, batch and loss (interface shown in Chinese).
Product screen: the client sees price, quantity and terms — never internal cost or margin (interface shown in Chinese).
Price date, tax basis and currency stated.
Fixed and variable loss handled separately.
Each packaging option costed separately.
Cost snapshot saved with the quote validity.
See the four layers first, then expand the method as needed. Formulas, BOMs and quotes share the same data.

Ingredient by ingredient, one tax basis and currency.
Usage from fill weight, order and batches.
Packaging loss and case allocation handled separately.
Manufacturing costs combined, exclusions listed.
In this example, 9,840 units fit in one batch; 9,890 need a second. Fixed loss occurs per batch, so cost does not scale linearly with order quantity.
Adjust order, loss and fee inputs to see the example cost breakdown and quote. This is for understanding the method; formal quotes need verified data and a frozen snapshot.
⛔ The first three go into cost C; the last goes into the price denominator. Why not multiply it into cost?Insurance depends on value → value depends on price → price includes insurance. It is circular, and only folding the rates into the denominator P = C ÷ (1 − m − Σr) solves it in one step. At a 25% margin, multiplying into cost undercharges by 8.3% at 5.3% total rates and by 10.8% at 10.3% — compare in the calculation chain below.
Missing key inputs prompt for entry, so unknowns are never treated as facts.
A price was entered without saying whether VAT is included. A 13% gap outweighs the whole loss debate, so the system never assumes pre-tax.
Net content by volume but no density: cleansing oil at 0.86 is overestimated by 16%, foundation at 1.35 underestimated by 26%. The direction is unknown, so 1:1 is never assumed.
Batch output would be negative. The formula still yields a number, but it means nothing physically.
Key details kept so sales and R&D can check them together.
F comes off before v. Writing (net usage + F) ÷ (1 − v) grosses up F as well — double-counting loss, and both versions still produce a number.
A blank loss rate is marked Not set and flagged, never silently zero. A zero-loss cost looks normal — it is just too low.
A silent zero makes an ingredient look free. The system lists unpriced lines and marks the cost as necessarily low.
Client-supplied items have no purchase price by design; flagging them would be a false alarm — and false alarms train users to ignore real ones.
Only scrapped over-issue of custom-printed cartons is costed; generic parts return to stock. Counting everything as loss would show a 90% loss rate on small-order labels.
Enter 96 units/case, not 0.010417. Switching to 48 units/case changes one integer and no BOM lines.
Whether to charge and how much the client should supply are separate questions. Under OEM contracts, the factory pays for any shortfall.
Quote lines carry cost and margin internally, but client documents never show them — printing them hands over your floor price.
Insurance, commission and platform fees are circular, so they are solved in closed form in the denominator. At a 25% margin and 10% rates, multiplying into cost undercharges by about 10.7%.
VAT is charged on top: pre-tax price × (1 + rate). Putting it in the denominator breaks both price and margin, so the system blocks it.
Fewer than 3 batches gets no fit. A reference F would end up in master data and skew extrapolation for every batch size.
Live and simulated values never reach a quote. Old prices cannot be rolled over after increases, and quotes expire on schedule.
A formula covers the bulk; a BOM covers bulk plus packaging. The same serum in a 30 ml glass bottle and a 50 ml plastic bottle means two BOMs, two costs and two quotes. On top of formula cost, a BOM adds packaging loss, minimum-pack over-issue, case-pack allocation, labor and tooling amortization, and finished-goods scrap.
Fixed loss (line and vessel residue) occurs once per batch regardless of size; variable loss scales with volume. Merge them into one percentage and small orders get quoted at a large-order loss rate — the smaller the order, the bigger the underestimate. Modeled separately, the system finds thresholds like 9,840 units = one batch, 9,890 units = two.
Yes. Enter input weight and output units from real production batches, and the system runs a least-squares fit across all of them to solve for fixed loss F and variable loss rate v. With fewer than 3 valid batches, or too little spread in batch size, it refuses to fit rather than offer a guess — solving from two points with ±2% measurement error can throw F off by 50%.
Rates charged on the selling price belong in the denominator: price = unit cost ÷ (1 − margin − total rates). They are circular (insurance depends on value, value on price, price includes insurance), so a closed-form solution is the only correct method. Multiplying them into cost instead, at a 25% margin, undercharges by 8.3% at 5.3% total rates and by 10.8% at 10.3%.
Client-supplied items are excluded from cost, but the system still calculates how many to request: net usage ÷ (1 − loss rate) × order quantity, rounded up to the minimum pack. For 10,000 units with 2% bottle loss, request 10,205; asking for 10,000 leaves you short, and OEM contracts usually make the factory pay for excess loss.
Quotes can only cite a frozen cost snapshot, which records the pricing date and every assumption. Quotes carry a validity period and expire on schedule; a new quote needs fresh prices and a new snapshot. Old quotes cannot be rolled over, and live or simulated values never go straight onto a quote.
Every unit price must declare its tax basis. If a price is entered without saying whether tax is included, the system refuses to compute cost and asks for it — it never assumes pre-tax. A 13% VAT gap is bigger than most loss debates; guess wrong and it eats your margin.
There are two modes. Simple mode needs just six inputs — net content, order quantity, bulk price, overall loss rate and so on — and suits small teams with limited data. As real batches and material master data build up, switch to the full layered-loss mode at any time without re-entering history.
Start with one formula and packaging option, and check prices, loss, batch size and quote terms.
Book a Demo