How to price a product so it actually makes money

A price makes money only once it covers every cost the order creates, not just the materials. In the worked example below a candle with £8.73 of materials — the number most sellers price from — actually costs £18.94 to make, sell, pack and post, so the £28.00 price that looks like a 68.8% margin is really a 32.4% one. Everything that follows is the arithmetic for closing that gap: the costs that only appear at checkout, the markup-versus-margin trap, how to price backwards from the margin you need, and why a 20% discount on that candle needs 2.56 times the orders just to stand still.

What a unit actually costs you

Almost every pricing method starts in the same place: add up what the thing cost to make, multiply by something, call it a price. The method is not wrong. The number fed into it usually is, because "what it cost to make" is only one of four cost blocks, and the other three only show up after the order exists.

The four blocks are: materials, the things consumed by the unit; direct labour, the time spent making and finishing it, valued at a rate you would actually accept; per-order costs, which appear at checkout rather than at the bench — the payment fee, the box, the tape, the label, the postage; and the failure allowance, the share of orders that break, go missing, get remade or get refunded, spread evenly across the orders that go right.

Here is one unit, fully counted. Every figure is illustrative and belongs to this example only — the point is the shape of the list, not the numbers in it, and you should replace all of them with your own.

Two of those lines deserve a second look, because they are the two people argue about.

The labour line. Leave it out and the same unit costs £14.74, and the £28.00 price appears to return £13.26, a 47.4% margin. That £4.20 did not vanish. It was paid, in hours, by whoever made the thing, and the only question is whether the price acknowledges it. At 100 orders a month, 18 minutes each, that is 30 hours of making — and if the price was set on materials alone, those 30 hours were worked for nothing and the business has no way of knowing it. A rate you would refuse from someone else is not a rate; put in a number you would actually be willing to be paid.

The line is also usually understated, because making is not the only time the order consumes. Add six minutes per order for packing, listing admin and buyer messages, at the same illustrative £14.00 an hour, and direct labour becomes £5.60 rather than £4.20. The unit cost rises to £20.34, the gross profit at £28.00 falls from £9.06 to £7.66, and the margin drops from 32.4% to 27.4%. Six minutes an order, five points of margin. That is the scale of thing this list is for.

The failure allowance. Work it out rather than guessing at it:

Failure allowance per unit = share of orders that go wrong × cost of putting one right

Putting one right here means making and sending another: materials £8.73 + labour £4.20 + packaging £1.15 + postage £3.49 = £17.57. If four orders in every hundred need that, the arithmetic is 0.04 × £17.57 = £0.70 per unit, every unit. A full refund with nothing returned lands in much the same place: you hand back the £28.00 you received and you have still spent the £17.57, so the loss is the £17.57 either way, give or take whatever the payment fee does. If your own rate is one in fifty rather than one in twenty-five, the allowance halves to £0.35 — use your figure, from your own records, not this one.

What this does to the headline number

Now compare the two ways of reading the same £28.00 sale.

• Against materials only: £28.00 − £8.73 = £19.27, which is 68.8% of the price. This is the number that gets quoted.
• Against the true unit cost: £28.00 − £18.94 = £9.06, which is 32.4%. This is the number that reaches the bank.

The familiar craft rule of thumb — price at three times your materials — would put this candle at £8.73 × 3 = £26.19. At that price the payment fee is a little smaller, so the unit cost is £18.91 and the gross profit is £7.28: a margin of 27.8%, not the 66.7% that "three times cost" sounds like it should mean. The doubling version is worse than disappointing. Two times materials is £17.46, against a unit cost of £18.78 at that price, which is a loss of £1.32 on every single order. A seller following that rule is paying customers £1.32 each to take the product, plus their own unpaid time, and the busier they get the faster it drains.

One thing the unit cost deliberately does not include: your fixed monthly costs. Rent, software, the accountant, your phone, the hours spent on things no order pays for. Those do not belong in the cost of a unit, because they do not change when you sell one more. They belong in the next question instead — how many units the gross profit has to cover them — and our break-even calculator answers that one directly.

Markup is not margin, and the difference is real money

These two words get used as if they mean the same thing. They are the same profit measured against two different denominators, and mixing them up is probably the single most expensive arithmetic error a small business makes, because it is invisible and it happens on every order.

Margin = (Price − Cost) ÷ Price
Markup = (Price − Cost) ÷ Cost

Take the £18.94 unit cost from the last section and hold it still for a moment. Add a 50% markup: £18.94 × 1.5 = £28.41. The gross profit is £9.47. Now measure it as a margin: £9.47 ÷ £28.41 = 33.3%.

A 50% markup is a 33.3% margin. Not approximately — exactly, always, for any cost, because the two are related by a fixed pair of conversions:

Margin = Markup ÷ (1 + Markup)
Markup = Margin ÷ (1 − Margin)

Which gives the conversion table. This is worth keeping somewhere you can see it, because the third column is where the money is: it shows the markup you would have had to apply to actually get the margin you thought you were getting.

Read the bottom row slowly. A seller who applies a 75% markup and believes they are running at a 75% margin is out by a factor of four in the work the price has to do: a genuine 75% margin needs a 300% markup, a price of £18.94 × 4 = £75.76 rather than £18.94 × 1.75 = £33.15.

What the confusion costs per order

Stay with the 50% case, because it is the most common one. The seller wanted 50p of every pound to be gross profit. To get that at a £18.94 cost, the price has to be £18.94 ÷ (1 − 0.50) = £37.88. They charged £28.41. The gap is £9.47 on every order, which is not a rounding error — it is the entire gross profit they did collect, missing a second time.

At 300 orders a year that is £9.47 × 300 = £2,841 of gross profit that was never priced in, on a business turning over £8,523 at the price actually charged. No customer was lost to get it. No extra unit was made. It is arithmetic that happened once, at the moment the price was set, and then repeated three hundred times.

The habit that prevents it is simple: state every profit target as a margin, never as a markup, because the margin is the one that answers the question you actually care about — what share of the money coming in stays with you. Markups are a way of getting to a price; margins are a way of judging one. If you want the conversion done for you in both directions, the margin calculator takes a cost and a price and gives you both numbers at once, along with the price any target margin would need.

One more distinction worth nailing down while we are here, because it catches people out at the end of the year. The margin in this section is a gross margin: price minus the cost of that unit. It is not your profit. Your fixed costs have not been paid yet, and neither have you if you have not been putting a wage in the labour line. A product with a healthy gross margin and a business with no net profit is an extremely ordinary combination, and it usually means the gross margin is real but there are not enough units passing through it.

Pricing backwards from the margin you need

Pricing forwards — cost, times something — means you find out what margin you have got after you have committed to the price. Pricing backwards means you choose the margin first and let the arithmetic tell you the price. It is the same equation rearranged, and it is a better habit because the thing you actually need to control sits on the left-hand side.

The one complication is that some costs are fixed per order and some are a percentage of the price, and a percentage cost moves when the price moves. Separate them and the equation solves cleanly:

Price = Fixed cost per unit ÷ (1 − target margin − percentage fees)

Where fixed cost per unit is everything that does not move with the price — materials, labour, packaging, postage, the failure allowance, plus any flat per-transaction fee — and percentage fees is everything charged as a percentage of the selling price, expressed as a decimal.

For our candle sold from the maker’s own website, the fixed block is materials £8.73 + labour £4.20 + packaging £1.15 + postage £3.49 + failure allowance £0.70 + the flat £0.25 part of the payment fee = £18.52. The percentage part is the 1.5% of the payment fee, so 0.015. Aim at a 40% margin:

£18.52 ÷ (1 − 0.015 − 0.40) = £18.52 ÷ 0.585 = £31.66

Check it the long way round, which is always worth doing the first time. At £31.66 the payment fee is 1.5% × £31.66 = £0.47, plus the flat £0.25. Total cost = £18.27 + £0.25 + £0.47 = £18.99. Gross profit = £31.66 − £18.99 = £12.67. And £12.67 ÷ £31.66 = 40.0%. The equation holds.

Run it at four targets and you have your own price ladder:

Notice what happens at the bottom of that table. Moving from a 30% target to a 60% target does not double the price — it takes it from £27.04 to £48.10, a rise of 77.9% — but it more than triples the gross profit per unit, from £8.11 to £28.86. That is the compression effect: near the top of the range, small price movements swing the profit hard, because the fixed cost block is being spread over a wider gap. It is also why the last few pounds of a price are the ones worth arguing about, and the first few are not.

Note the £38.19 in the 50% row. In the previous section, holding the cost still at £18.94, we said a 50% margin needed £37.88. The 31p difference is the payment fee growing with the price. It is small here because the percentage fee is small; on a channel taking ten times as much it is not small at all, which is the next section.

Choosing the target instead of guessing at it

The target margin is not a preference, it is a requirement, and it comes from your fixed costs. Work out what the business has to pay every month regardless of sales — rent, software, insurance, the accountant, your own non-production time — and then ask how many units each candidate price needs to sell to cover it.

Say fixed costs are £950 a month, an illustrative figure. At the 40% price the gross profit per unit is £12.67, so:

£950 ÷ £12.67 = 74.98 → 75 units a month before a penny of net profit

At the 30% price the gross profit is £8.11, so the same £950 needs £950 ÷ £8.11 = 117.1, call it 118 units — 57% more orders, 57% more packing, 57% more of everything, to reach the same standing start. At the 50% price it needs £950 ÷ £19.10 = 49.7, call it 50 units. Those three numbers, not a feeling about what the market will bear, are what tells you whether a target margin is the right one: pick the target whose unit count you can realistically sell. The break-even calculator does this sum for a whole business, with your fixed costs and your own margin in it.

Two practical cautions before you fix a number. First, round the output up, never down. The equation gives £31.66; charging £31.95 or £32.00 costs you no customer who was willing to pay £31.66 and adds pure gross profit, whereas rounding down to £29.99 quietly gives away £1.67 an order, which at 300 orders is £501. Second, if you are VAT registered or approaching registration, every price on this page becomes a different number, because the sum itself changes shape. That is a conversation to have with your accountant before you set the ladder, not after.

The same product on a marketplace and on your own site

The backwards equation does not change when you switch channel. The numbers you feed it do, and they change enough that the same product, at the same margin, is a materially different price in two places.

A marketplace adds a commission charged as a percentage of the order, and usually one or more flat fees on top. The real rates are published by each platform and you should take them from the platform itself rather than from a page like this one — for a worked example using real published figures, see what Etsy takes from a sale. For the arithmetic here, assume an illustrative marketplace charging 10% of the order plus a flat £0.20 listing fee, on top of the same 1.5% + £0.25 payment processing.

That makes the percentage block 10% + 1.5% = 11.5%, and the fixed block £18.52 + £0.20 = £18.72. At the same 40% target:

£18.72 ÷ (1 − 0.115 − 0.40) = £18.72 ÷ 0.485 = £38.60

The same candle, the same margin, and the price has gone from £31.66 to £38.60 — 21.9% higher — purely to carry the commission. Put side by side:

The bottom three rows are the ones that get left out of this comparison, and leaving them out is what makes "the marketplace takes too much, I should sell direct" sound obviously true when it often is not. The commission is not a pure tax. It is the price of footfall. On your own site there is no commission and no footfall either, and whatever you spend to replace it — ads, content, time, a subscription — is a real cost of every order, it just arrives on a separate invoice where the unit economics never see it.

The honest version of the sum is the one above: divide your monthly cost of getting visitors by the orders it produced, and put the answer in the unit cost. At an illustrative £400 a month producing 90 orders that is £4.44 an order, and the direct channel that looked like a 40% margin is a 26.0% one — worse, on these illustrative numbers, than the marketplace it was supposed to beat. Change either figure and the answer flips: £400 producing 200 orders is £2.00 an order and direct wins comfortably. Neither channel is structurally better. The arithmetic decides, and it decides differently for different businesses.

When you cannot show two prices

Sometimes you have to display the same price in both places, whether because a platform expects it or because customers will simply see both. In that case the marketplace is the channel that sets your price everywhere, because it is the one with the higher cost to serve. Price for it, and your own site becomes the better order rather than the cheaper one.

At £38.60 on the maker’s own site: percentage fee 1.5% × £38.60 = £0.58, plus £0.25 flat, on the £18.27 base gives a cost of £19.10 and a gross profit of £19.50 — a margin of 50.5% instead of 40.0%. Every order that comes direct is worth £4.06 more than the identical order through the marketplace (£19.50 against £15.44), which is a concrete reason to put a card in the box asking people to come back directly, and a concrete number to judge any loyalty discount against.

Two further things move this sum and both are worth checking before you settle a channel price. Marketplace advertising, where it exists, is charged on top of the commission and on some platforms is not optional above a certain size — add its percentage to the 11.5% block and re-run the equation, and you will see the required price jump again. And postage rules differ by channel: if one charges commission on the postage you collect and the other does not, the two are not comparable at all until that is in the arithmetic.

When to price on value instead of cost

Everything so far has worked forwards or backwards from cost. Cost-based pricing has one enormous virtue — it cannot produce a price that loses money — and one serious limitation: it has no opinion whatsoever about what the thing is worth to the person buying it. Your costs are a fact about you. The price ceiling is a fact about them.

That matters because for some products the two numbers are nowhere near each other. Consider a digital template or a small software tool where the cost to produce and support one sale is £42, an illustrative figure covering support time and fees. Cost-plus at a 100% markup gives £84 and a 50% margin, and that feels responsible. Now look at the buyer’s side: if it saves them six hours a month and their own time is worth £30 an hour, the thing is generating £180 a month of value for them. Priced at £120 it returns a margin of (£120 − £42) ÷ £120 = 65%, and the buyer still keeps £60 of the £180 in the first month alone. Same cost, same product, the margin moved 15 points, and nobody was overcharged — the price was simply set against the value rather than against the bench.

The four conditions

Value pricing is not a decision you can make on your own behalf. It works when four things are true, and you can check each of them without guessing:

1. The buyer’s gain is expressible in their own numbers. Hours saved times their hourly cost. Waste avoided. Revenue enabled. If you cannot write the buyer’s benefit as an arithmetic sum in their units, you do not have a value price, you have a hopeful one.
2. The thing is hard to compare directly. If an identical item is listed next to yours with a visible price, the buyer prices it for you. Value pricing needs some genuine difference — a bundle, an outcome, a guaranteed turnaround, an audience, a format nobody else offers — that makes a side-by-side comparison inexact.
3. Your cost to serve does not rise with the value you deliver. This is the one that makes value pricing worth the trouble. If the buyer who gets £2,000 of value costs you the same to serve as the one who gets £200, the extra price is nearly all margin. If serving the high-value buyer costs proportionally more, you are back to cost-plus with extra steps.
4. You can tell the two buyers apart before you quote. Different sizes, different use cases, different volumes. If every buyer looks identical at the point of sale you cannot charge them differently, and you are left setting one price for the middle of the range.

Fail any of those four and cost is still the better instrument. In particular, a commodity sold on a channel where buyers sort by price is a cost problem, not a pricing problem: the lever is the cost line, not the price line, and the only durable fix is to make the unit cheaper to produce or to stop selling something that can be compared that precisely.

Cost still sets the floor

Value pricing sets the ceiling. It never replaces the floor. Whatever the buyer’s world says a thing is worth, a price below your true cost per unit loses money on every order, and the losses grow exactly in proportion to how well the product sells. Recall the "two times materials" candle from the first section, losing £1.32 an order: at 100 orders a month it loses £132, at 400 orders it loses £528, and the seller experiences the second month as a busier and more successful one. Volume never rescues a price below cost. It only speeds it up.

Which is also the practical way to test a value price without betting the business on it. Set the floor with the backwards equation, set the ceiling with the buyer’s arithmetic, and if there is daylight between them, move the price up in small steps and watch what happens to total gross profit, not to units sold. Units are supposed to fall when a price rises. The question is only whether they fall by more or less than the margin gained — which is the same arithmetic as a discount, run in reverse, and it is the last section of this page.

The discount trap: the volume a discount has to find

A discount is not a marketing decision with a financial side effect. It is a direct transfer out of gross profit, and because it comes off the price while the cost stays exactly where it was, it takes a far bigger bite than its headline suggests. A 20% discount does not cost you 20% of anything you care about. It can easily cost you 60% of your gross profit.

The arithmetic is short enough to do in full. Let the price be P, the gross margin be m as a fraction of the price, and the discount be d, also as a fraction of the price. The gross profit per unit before the discount is P × m. The discounted price is P × (1 − d), and the cost has not changed, so the new gross profit per unit is P(1 − d) − P(1 − m) = P × (m − d). To make the same total gross profit from N units as you used to make from the units you were selling:

New volume ÷ old volume = m ÷ (m − d)
Extra volume needed = d ÷ (m − d)

The discount comes straight off the margin, one point for one point. That is the whole mechanism, and it is why the answers get violent so quickly: the denominator is not your margin, it is your margin after the discount has eaten into it.

The table

Extra unit volume needed simply to make the same gross profit as before. Read down to your margin, across to the discount you are considering.

"Impossible" is literal, not rhetorical. At a 20% margin, a 20% discount sells the unit for exactly what it cost: the gross profit per unit is zero, and no quantity multiplied by zero has ever covered a rent bill. Anything past that line loses money per unit, so every additional order makes the month worse. There is no volume that fixes it.

The table also explains something that looks unfair from the outside. A business at a 60% margin can run a 20% sale and needs 50% more units to break even on it. A business at a 25% margin running the same 20% sale needs five times as many. Low-margin businesses cannot use discounting as a tool at all, which is precisely the position discounting tends to put them in.

Our candle, exactly

The formula above treats the unit cost as fixed. In real life the percentage-based part of the cost falls slightly when the price falls, so the true answer is a shade kinder than the table. Take the candle at £28.00, a 32.4% margin, and knock 20% off:

£9.06 ÷ £3.54 = 2.56, so 100 orders have to become 256 orders — an increase of 156% — for the month to end in exactly the same place. The simplified formula would have said d ÷ (m − d) = 20 ÷ (32.4 − 20) = +162%, so the shortcut was six points pessimistic, which tells you the shortcut is good enough to make decisions with.

And "the same place" is generous to the discount, because those 256 orders are not free. They are 2.56 times the making hours, 2.56 times the packing, 2.56 times the postage queue, 2.56 times the stock bought in advance and 2.56 times the exposure to breakage and returns — and the hours in that list were costed at an illustrative £14.00 an hour, so they are your hours. The business is two and a half times busier, carries two and a half times the risk, holds more cash in stock, and banks the identical amount. Run at those levels, a discount is a way of buying work.

The same sum in reverse

Price rises obey the identical mechanism, which is the more useful half for most sellers. If you raise the price by r as a fraction of the old price and lose some customers, the volume you can afford to lose before you are worse off is:

Volume you can afford to lose = r ÷ (m + r)

At the candle’s 32.4% margin, a 10% rise means 10 ÷ 42.4 = 23.6%. Nearly a quarter of the customers can walk away and the month ends no worse. Exactly, allowing for the payment fee rising with the price: at £30.80 the gross profit per unit is £11.82, and £9.06 ÷ £11.82 = 0.7665, so you can lose 23.3% of the orders and stand still. Most businesses do not lose anything like a quarter of their customers to a 10% rise, which is the whole reason this sum is worth doing before the next one is assumed to be impossible.

None of this says never discount. It says price the discount first. If a promotion has a job — clearing stock that is costing you space, acquiring a customer you can reasonably expect to buy again, filling a genuinely dead week — then run the table, decide what extra volume you would need, and judge whether that number is plausible before the offer goes out rather than after. Every question on this page reduces to the same discipline: know your true cost per unit, express the target as a margin, and check the arithmetic before the price is public rather than at the end of the year when it is a fact.

Arithmetic and general information only — not financial, tax, legal or investment advice. Your own figures, and your own accountant, decide what any of this means for you.

Frequently asked questions

How do I work out the true cost of a product?

Add four blocks, not one. Materials consumed by the unit; direct labour at a rate you would actually accept, covering making, packing and admin; per-order costs that only appear at checkout, which are the payment fee, packaging and postage; and a failure allowance, being the share of orders that go wrong multiplied by the cost of putting one right. In the worked example on this page those four blocks turn 8.73 pounds of materials into a true cost of 18.94 pounds per unit.

Is a 50 percent markup the same as a 50 percent margin?

No. A 50 percent markup is a 33.3 percent margin, because markup is measured against cost and margin against price. Margin = markup divided by (1 + markup). To actually reach a 50 percent margin you need a 100 percent markup. On an 18.94 pound cost that is the difference between charging 28.41 and charging 37.88, or 9.47 pounds on every single order.

What is the formula for pricing from a target margin?

Price = fixed cost per unit divided by (1 minus target margin minus percentage fees). Fixed cost per unit is everything that does not move with price, including any flat transaction fee. Percentage fees are anything charged as a share of the selling price, as a decimal. With a fixed block of 18.52 pounds, 1.5 percent fees and a 40 percent target, the price is 18.52 divided by 0.585, which is 31.66.

Should I charge more on a marketplace than on my own website?

The margin arithmetic says yes, because the commission is a percentage cost your own site does not carry. In the illustrative example here, holding the margin at 40 percent, the marketplace price is 38.60 against 31.66 direct. But your own site is not free either: divide what you spend each month attracting visitors by the orders it produces and put that figure in the unit cost, and the direct channel can easily end up the thinner of the two.

How much extra do I need to sell to cover a discount?

Extra volume needed = discount divided by (margin minus discount), with both as percentages of price, because the discount comes off the margin point for point. At a 30 percent margin, 10 percent off needs 50 percent more units and 20 percent off needs 200 percent more. At a 20 percent margin, a 20 percent discount sells at cost and no volume can recover it. The same formula in reverse shows how much volume a price rise can afford to lose: rise divided by (margin plus rise).

Other guides

• How much does Etsy take from a sale?
• How to build a cash flow forecast
• Agency utilisation rate: the formula, and the denominator that decides it