Cost-plus pricing: the formula, a calculator, and prices tied to cost

Cost-plus pricing sets a selling price by adding a markup to what one unit costs you: its materials, its labour and a share of overhead. The formula is price = unit cost × (1 + markup). Manufacturers, distributors and retailers often use it because it is a fast and straightforward way to set prices. Its weak points are a unit cost that has gone out of date and a markup that ignores what buyers would pay. This guide explains the cost build-up, markup versus margin, when cost-plus pricing works and when it leaves money on the table, and how AI keeps cost-based prices tied to the costs in your ERP. The calculator works out price, profit and margin, and the markup you need for a target margin.

Derik Lawlis Founder, ThriveAI

The cost-plus pricing formula

Cost-plus pricing needs two inputs, the cost of one unit and a markup percentage, and produces a selling price.

BDC, the Business Development Bank of Canada, defines the two percentages in its guide to setting prices: “The markup is expressed as a percentage of cost of goods sold or cost of sales,” while “Gross margin is a related term and is gross profit as a percentage of revenues.” Markup is measured on cost and margin on price, so the same dollar of profit is a larger percentage as a markup than as a margin.

Take a part that costs $42.00 to make. A 30% markup adds $12.60, for a selling price of $54.60. That $12.60 is 23.1% of the price, so a 30% markup gives a 23.1% margin. BDC’s own example shows how far apart the two numbers can be: an item that costs $3.50 and sells for $10.00 carries a markup of 186%, which works out to a gross margin of 65%.

Cost-plus pricing calculator

Enter your unit cost, either as its parts or as one total, and a markup. The calculator shows the selling price, the profit on each unit and the margin that markup gives you. The last part works backwards, from the margin you want to the markup you need. It starts with the example used on this page.

Markup versus margin

Markup is profit as a share of cost. Margin is profit as a share of the selling price. Whenever there is a profit, the margin is the smaller of the two percentages, and they convert with two formulas: margin = markup ÷ (1 + markup), and markup = margin ÷ (1 − margin).

A costly mistake is to treat a target margin as a markup. On goods that cost $700,000, a 30% markup brings in $910,000 of sales and $210,000 of gross profit, a margin of 23.1%. Pricing the same goods for a 30% margin brings in $1,000,000 of sales and $300,000 of gross profit. If your budget assumes a margin and your price list applies a markup, check which one each number means.

Building the unit cost

A unit cost has to include everything that one unit consumes. BDC’s guide starts there: “Direct costs include the cost of raw materials, including duty, freight or shipping charges, plus direct labour costs that are incurred to produce a product or provide a service.” Overhead, the cost of running the plant or the warehouse, is then spread across the units made or sold.

Here is the cost build-up for the $42.00 part used on this page. The numbers are made up to show the method.

The overhead rate is where judgment comes in. OpenStax’s managerial accounting text explains that “A predetermined overhead rate is calculated at the start of the accounting period by dividing the estimated manufacturing overhead by the estimated activity base,” such as labour hours or machine hours. In the example, $1.8 million of budgeted overhead spread over 20,000 machine hours gives $90.00 an hour. The same text notes that this traditional method “works most effectively when direct labor is a dominant component in production,” and that many organizations have moved to activity-based costing, which assigns overhead by the activities each product uses, such as machine setups.

A distributor’s unit cost is a landed cost: the supplier’s price plus inbound freight, duty, brokerage and handling, spread over the units received. A food producer’s cost starts from the recipe and allows for yield, because trimming and cooking losses mean more raw material goes in than comes out as product. In every case, use the price you paid on the latest receipts. A standard cost set at the start of the year can be months behind.

When cost-plus pricing works

BDC’s overview of common pricing strategies notes that “Retailers, manufacturers, restaurants, distributors and other intermediaries often think of cost-plus pricing as a fast and straightforward way of setting prices,” and that “Many entrepreneurs and customers think cost-plus pricing is the only way to set prices.” It fits some situations well:

• Made-to-order work. A custom part or a one-off job has no market price to compare with, and its cost can be estimated from the drawing and the routing.
• Long lists of low-value items. BDC’s example is a hardware store, where customers see most items, “such as nuts, bolts and washers,” as low value, so a markup by category is simpler than pricing each item.
• Open-book agreements. Some customers pay cost plus an agreed fee and audit the costs, which makes the cost build-up the price.
• Volatile inputs. When most of the cost is a commodity such as steel, copper or resin, a price tied to cost moves with the market for that input.

When it leaves money on the table

BDC names the main drawback plainly: cost-plus pricing “doesn’t consider the customer. If you use this pricing strategy, you may lose potential profit.” In the same hardware store, customers “will see value in some items, such as electric tools or air compressors,” which can be priced on value instead. The same thing happens at a manufacturer when a part is urgent, hard to source elsewhere, or engineered for one customer.

It can also price you out. In BDC’s guide to setting prices, a BDC business consultant warns that you cannot choose a markup on math alone: “If your competitors all have lower profit margins and you offer a higher price that would give you a higher margin, you could lose sales.” Pricing intelligence covers tracking what competitors charge.

When the cost underneath is stale

Costs also move between price reviews. In August 2026, Statistics Canada reported that its Industrial Product Price Index was 13.5% higher than a year earlier, the 23rd consecutive month of year-over-year increases, and that its Raw Materials Price Index was up 22.8%. Over the same year, unwrought copper rose 49.3% and unwrought aluminum 27.6%, and plastic resins rose 10.7% in August alone. A price list set once a year trails costs like those for most of the year.

When volume falls

Overhead per unit depends on volume. If the plant in the example runs 15,000 machine hours instead of 20,000, the same $1.8 million of overhead becomes $120.00 an hour, the unit cost rises from $42.00 to $49.50, and a 30% markup asks $64.35 for the same part. Raising prices because volume fell can push more volume away, so price on the overhead rate at normal volume and review the idle hours separately.

Cost plus in contracts

In a contract, a fee that grows with cost gives the supplier no reason to cut cost. The Government of Canada’s buyer’s guide says a basis of payment that pays “actual costs incurred plus a fixed percentage fee” is not recommended “because it provides little or no costs control and actually encourages contractors to raise costs to increase profit.” US federal rules go further: FAR 16.102 says “The cost-plus-a-percentage-of-cost system of contracting shall not be used,” and the cost-plus-fixed-fee contract described in FAR 16.306 pays a fee fixed at the start, which still provides the contractor “only a minimum incentive to control costs.” If you sell under an open-book or government contract, read its definition of cost before you set a price.

How AI keeps cost-based prices tied to live costs

Most of the effort in cost-plus pricing goes into keeping unit costs current across hundreds or thousands of items. AI changes that work in five places:

• Reading cost changes as they arrive. Purchase receipts show the price you actually paid. Supplier price-increase letters, quotes and freight bills arrive as PDFs and emails, and language models read them into structured costs that a buyer confirms. Text extraction from images covers reading those documents, and AI in procurement covers the purchasing side.
• Rolling costs up. A change in one raw material flows through every bill of materials or recipe that uses it, so one resin increase updates the cost of every product made from that resin.
• Checking every price against its floor. For each item and each customer-specific price, the software compares the current price with the unit cost plus your minimum markup, and lists the ones below it with the gap in dollars.
• Drafting the new prices. It proposes a new price list or quote with the cost trail behind each line, rounded to your price points, and a person approves it before anything changes in the ERP or reaches a customer.
• Checking the market before you send. A cost-based price can sit far from what competitors charge. Price optimization covers setting the markup from quote wins and losses, and margin analysis covers what each product and customer earns.

For example, when a food producer’s packaging film goes up 8%, the software lists which of its 300 products now earn less than the 25% minimum markup, by how much, and the price each would need to get back to the minimum.

What data it needs

• Item master. Every item with its unit of measure, category and current price.
• Bills of materials or recipes. The quantity of each input per unit, with scrap or yield allowances.
• Routings and labour standards. Hours per operation and the machine or work centre used.
• Purchase receipts and supplier price lists. The price actually paid, with dates, and any increases already announced.
• Freight, duty and brokerage. The inbound bills that belong in landed cost.
• Overhead budget and hours. The totals behind each overhead rate.
• Price lists and customer prices. List prices, customer-specific prices, contracts with fixed prices, and recent quotes.

These records live in the ERP and the accounting system, plus spreadsheets for whatever the ERP does not hold. AI for ERP covers the ways to read them.

How to start

1. Pick the items that matter. Start with the 50 to 100 items that bring in most of your revenue, plus any whose main input moved more than 10% in the past year.
2. Rebuild their costs from receipts. Compare the cost behind each current price with a cost built from the latest purchase prices, and list the gaps.
3. Write the markup rules down. Minimum markup by category or customer group, rounding rules, and who approves exceptions.
4. Let the software draft, and approve each change. Human in the loop covers designing an approval step people actually read.
5. Repeat every month. Once the cost roll-up reads the ERP directly, the same check can run monthly instead of once a year.

Risks and limits

• Out-of-date inputs. If bills of materials or routings are out of date, every price built on them is too. Fix the inputs for the items you start with.
• Contract terms. Some customer agreements fix prices for a term or require notice before an increase. The software has to know which prices it may change.
• Overhead method. How you spread overhead changes every unit cost, so agree on the method before you automate it.
• Approval before publishing. Keep a named approver for every price change, and send nothing to a customer as a draft.
• Prices set independently. Set your markups yourself. Agreeing on prices with a competitor is an offence under section 45 of the Competition Act.

How ThriveAI helps

ThriveAI is an AI engineering company in Ottawa that builds private AI systems on a company’s own data, for businesses that make, move or sell physical goods. For pricing, that means reading the costs in your ERP, purchase receipts and supplier documents, rolling them up through your bills of materials or recipes, and drafting price changes for a named person to approve, with the cost trail behind every line. Every answer shows where it came from.

ThriveAI’s systems read your ERP as it is, including older versions installed on your own server, and every connection only reads data. They are designed to keep each client’s data on its own server in Canada. You choose a model on that server or a hosted model under a written zero data retention agreement, and a hosted model may process requests outside Canada. Derik Lawlis, the founder, leads every project and stays close to the build. About ThriveAI covers the company.