SpotPeaks

Setting a price that pays for its ads

One formula that works back from what a customer costs to buy, with US duty, card fees and returns in it.

12 min read
By the end you can
  • Price from the ad cost, not a markup rule
  • Put duty, card fees and returns into the price
  • Tell when a product is too cheap to advertise

The short answer

Do not multiply your product cost by three and call it a price. Work backwards from what a customer costs you to buy, because that is the biggest number in the whole calculation and it barely moves when your price does. The formula is:

Price = (landed cost + duty + fixed card fee + target ad cost per sale) divided by (1 minus card percentage minus return rate minus target net margin)

Everything on this page is that one line, unpacked. It takes about five minutes with a calculator, it uses only numbers you can know before you spend anything on ads, and it answers the question markup rules cannot: whether the price you are about to set can survive the ad auction you are about to enter.

Why markup rules point the wrong way

The 3x rule sets your price from your cost. That feels right and it is backwards, because your cost is the small, controllable number and the cost of finding a customer is the large, mostly uncontrollable one.

Here is the pivot the rule misses. Cost per purchase is roughly fixed per order. Facebook does not charge you more to sell a $90 product than a $25 one, so raising your price adds almost pure headroom, while raising your cost only shifts a line that was never the problem. Triple Whale’s Meta benchmark set, which we fetched directly and which covers nearly 35,000 brands over 1 January to 31 December 2025 (page last updated 7 April 2026), puts the median cost per acquisition at $38.19, with a spread from $29.99 in lifestyle and boutique to $49.48 in electronics. Those are commercial aggregates rather than an audited census, so read the direction and ignore the decimal places.

A $25 product at a 3x markup is a perfectly obedient application of the rule and it cannot pay a $38.19 bill. No amount of store optimisation fixes that, which is why a store with traffic and no sales is often a store whose price was wrong before the first visitor arrived.

Every term in the formula, and where to get it

  • Landed cost. What the unit and its shipping to your customer actually cost you, not the supplier headline. Take it from a real quote for the variant you will sell.
  • Duty. US import duty, which every dropshipped order now owes: duty-free entry for parcels under $800 ended for goods from China on 2 May 2025 and for every country on 29 August 2025. It is a percentage of what you pay for the item, not the shipping, so turn it into dollars per order before it goes in. Zero per order if the product ships from US stock, because the importer already paid it and it sits inside the supplier’s price. More on this below.
  • Fixed card fee. The per-transaction part, the $0.30, not the percentage. The percentage belongs in the divisor because it scales with price.
  • Target ad cost per sale. The one judgement call. Use the benchmark until your own account has spent a few hundred dollars, then replace it with your real number and re-run the calculation.
  • Card percentage. Shopify Payments is 2.9% + $0.30 on Basic in the US. Take yours from your own admin rather than from anyone’s table, including ours.
  • Return rate. A refund costs you the whole sale price, not your margin on it, so every point of return rate costs one percent of your price. The returns guide derives that identity in full. Use your category band until you have your own.
  • Target net margin. What you want left over. 15% is a reasonable working figure and the honest one to plan with.

Worked example: what price does the median ad cost demand?

A $12 landed cost, US card fees at 2.9% + $0.30, a 10% return rate, a 15% target net margin. The divisor is 1 minus 0.029 minus 0.10 minus 0.15, which is 0.721. Turn the formula around first and ask what each price level can afford to pay for a customer, since that is the number your ads manager will report back to you.

Your priceMost you can pay per saleVerdict against the $38.19 median
$25$5.73Needs to beat the median by about 7x
$35$12.94Needs to beat it by about 3x
$50$23.75Needs clearly better than median advertising
$70$38.17Median advertising is exactly enough
$90$52.59Room to be worse than median and still profit
$120$74.22Room, but a harder sale to a cold audience

The $70 row is the interesting one. Run the formula forwards at the median cost per acquisition and it returns $70.03, and the median AOV reported in that same Triple Whale set is $71.69. Two numbers landing in the same place is worth noticing, but be careful about what it proves: both come from one publisher, and that publisher’s median ROAS of 1.86 already encodes roughly the same ratio between order value and acquisition cost. So this is the dataset agreeing with itself, not independent confirmation.

What it does suggest is worth having anyway. The typical order value in ecommerce is not mostly a matter of taste. It sits near the level the ad market forces, because stores priced far below it stop buying ads and drop out of the dataset. When you price a $25 impulse product, you are not choosing a cheaper positioning, you are choosing to be an outlier against an auction that has already set a floor.

Your creative sets your price more than your supplier does

Same store, same $12 product, same fees. The only thing that changes is how good the advertising is, and the price required moves by nearly $55.

Ad cost per sale you can achievePrice required for a 15% net marginWhat that level of advertising means
$10$30.93Excellent creative, or organic traffic doing the work
$15$37.86Well above median, achievable and not typical
$25$51.73Better than most advertisers manage
$38.19$70.03The Triple Whale median across nearly 35,000 brands
$49.48$85.69The electronics median in the same set

Read that as an instruction rather than a table. Before you have any creative, you have no evidence you will beat the median, so price the third or fourth row. If you price the top row on day one you have quietly bet your whole budget on being an excellent advertiser before you have made a single video. You can always come down later; coming down is a normal promotion, while going up asks every returning visitor to accept a price rise. Model your own version with the pricing calculator, then sanity-check the result against the breakeven ROAS calculator and the guide behind it.

US import duty costs more than its rate, and the reason matters

Duty-free entry for low-value parcels ended for goods from China on 2 May 2025 and for every country on 29 August 2025. Customs made the suspension indefinite in June 2026 and the law ends it for good on 1 July 2027, so every order a supplier ships to your customer from abroad is an import that owes duty. As of October 2026 a China-made product pays three layers: the base duty for its tariff code (0% to about 6% for most consumer goods, far higher for clothing and shoes), the Section 301 list it sits on (7.5% on List 4A, where most consumer goods are, and 25% on Lists 1 to 3, which cover machinery, electronics, furniture and many household goods), and a 12.5% Section 301 layer on China-origin goods since 24 July 2026, which leaves out steel, aluminium and the other Section 232 goods. So a List 4A product with no base duty pays 20%, a List 3 product 37.5% and an unlisted one 12.5%, each plus its base duty.

The rate for your exact product comes from its HTS code at hts.usitc.gov or from a customs broker, and the US landed cost calculator walks through each layer. Price it as if you pay it, which is what a supplier’s “import charges included” option does. The alternative, the carrier billing your customer at the door, comes back as refused parcels, refunds and chargebacks.

Now put it through the formula. Take the same $12 landed cost, of which $9 is the item and $3 the shipping, the same US card fees, return rate and target margin (a divisor of 0.721), and the $38.19 median ad cost per sale. Duty is charged on the $9, not the $12.

Where it ships fromDuty per orderPrice required
No duty line, as in the tables above$0$70.03
China, not on a Section 301 list (12.5%)$1.13$71.60
China, List 4A with no base duty (20%)$1.80$72.52
China, Lists 1 to 3 with no base duty (37.5%)$3.38$74.72

Each dollar of duty raises the price you need by about $1.39, not by $1. A cost added to an order has to carry its own share of the card percentage, the return risk and the margin you wanted, so it arrives at the price line multiplied by 1 divided by 0.721. The same multiplier applies to anything else you add per order: a $2 insert card is $2.77 of price, a $4 shipping upgrade is $5.55.

Because the duty is a percentage of the item, it grows with your product cost, and bundling does not dodge it: three units owe three times the duty of one. A product from US stock carries no duty line at all, which is part of what its higher supplier price is paying for, along with the shorter delivery that keeps returns down. Every product in SpotPeaks is priced for a US store this way already, with the item, shipping, an estimate of the US import duty, the payment fee and an estimated ad cost per sale in the numbers, so you start from the full cost rather than the supplier headline.

Should the price end in 9?

There is one properly designed piece of evidence here and a great deal of folklore. Anderson and Simester published three field experiments in Quantitative Marketing and Economics in 2003 in which price endings were manipulated for real customers of a real retailer. A $9 ending increased demand in all three, and the effect was stronger for new items than for items the retailer had sold in previous years. The paper’s abstract as indexed also reports that $9 endings were less effective when the retailer used “Sale” cues, though we could not open the full paper in this session to verify that line.

The widely circulated claim that charm pricing lifts sales by 24% is not from that study, we could not trace it to any inspectable source, and the pricing literature that reviews the whole body of work reports a mix of positive, null and occasionally negative results. So use the 9 ending, because the direction has real experimental support and it costs you a cent, and do not build a business case on a percentage nobody can show you. The new-item finding is the genuinely useful part for a dropshipper: your product is new to every visitor you have, which is exactly the condition where the effect was strongest.

One caveat worth keeping. The same literature notes that a 9 ending reads as value rather than quality, which is why premium brands round. If you are selling a $90 item on perceived quality, $90 may serve you better than $89.99.

Charge one number, including shipping

Baymard Institute’s cart-abandonment meta-analysis, which we fetched directly and which draws on 50 studies (last updated 22 September 2025), puts the documented average abandonment rate at 70.22%. Excluding people who were only browsing, the top reason is extra costs being too high at 40%, ahead of slow delivery at 20% and payment-security worries at 19%. Note that the figure circulating widely in 2026 roundups is 48%, which is not what Baymard’s own page currently says, so we use theirs.

For pricing, that single row is the whole argument for building shipping into the price and showing one number. The customer who sees $69.99 and then $6.90 of shipping at checkout has been given two chances to say no. The one who sees $76.89 has been given one. Your formula does not care which way you split it, because the landed cost line already includes shipping to the customer.

Free-shipping thresholds are the usual follow-up question, and the honest answer is that the mechanism is real and the published numbers are not usable. The circulating claims of a 12% to 24% lift in average order value, or 15% to 25%, or 17% to 30% with a progress bar, all come from ecommerce app vendors’ own resource pages, none publish a methodology, and they disagree with each other by more than the effect they describe. Take the direction, run your own threshold, and measure it yourself.

What our ad data can and cannot tell you about price

Be straight about this: our radar tracks which products are being advertised and how long those ads have run. It does not track selling prices, so we cannot hand you a price band per category, and you should be suspicious of anyone who does. What you can do is open the listings behind a niche and read the prices yourself, which takes ten minutes and beats any table.

The SpotPeaks radar currently tracks 3,254 products with ads on our radar across 124 niches.

Three niches where the pricing arithmetic above plays out very differently:

Wireless earbuds is the deepest bench we track and the hardest pricing problem on this list, because electronics carries the highest cost per acquisition in the Triple Whale set at $49.48. The table above says that demands a price around $85.69, and the market price for dropshipped earbuds is nowhere near it. That gap, not the competition, is why cheap earbuds consume so many first budgets. Kitchen gadgets is the classic $15 to $25 impulse category, which is the exact band the $25 row kills, so the winners there are almost always sold as multi-piece sets rather than as one gadget. Home decor loses gross margin to bulky shipping, but it upsells naturally into several units of the same product, which makes a bigger order out of a single ad cost.

As of August 2026 the widest niche on our radar carries 49 products being advertised at once, which is a small enough field to read every listing before you set your own number. Some of the ads on our radar have run for months (counted from the platform’s start date, so not proof that an ad ran without a break). A price whose ad keeps running through months of rising ad costs is the only public evidence anyone has that a given price point in a given category can carry a permanent ad bill. Browse what is running by niche on Facebook or across the whole winning-products radar, and check how crowded a category is with the saturation checker before you commit, because a crowded auction has usually already set your ceiling.

What to do next

In order, and none of it needs an ad budget:

  • Get a real landed cost for the exact variant, including shipping to the customer.
  • Add US import duty: the rate for its tariff code times what you pay for the item, unless it ships from US stock.
  • Pick a target ad cost per sale you can defend. Until your account has data, use a benchmark and assume you are not yet better than it.
  • Run the formula, then check the answer with the profit margin calculator and the profit simulator.
  • End it in 9. US prices are normally shown before sales tax, which is added at checkout, so the number the formula gives you is your sticker.
  • Leave the compare-at field empty until you have a real former price to show. An invented “was” price misleads every buyer who sees it, and the EU has a specific rule against it, covered below.
  • If the price the formula returns is one nobody would pay for this product, you have not found a pricing problem. You have found the wrong product, and validating the next one is cheaper than discounting this one.

The honest limit, stated plainly: no formula can guarantee you profit, and this one cannot either. It contains one estimate, your target ad cost per sale, and how your ads actually perform with real money live is genuinely in your hands. What the arithmetic does is stop you launching a price that could never have worked, which is the quietest and most common way a first budget disappears.

If you also sell into Europe: the €3 duty, VAT and discount rules

Everything above prices a US store. Selling to EU consumers changes three lines of the calculation: a flat customs duty in place of the US percentage, a sticker that has to include VAT, and a legal rule on compare-at prices.

The €3 duty costs €4.11, and the reason matters

Since 1 July 2026 the EU charges a flat €3 customs duty on low-value parcels arriving from outside it. We re-verified this against the European Commission’s own announcement of 29 June 2026: goods worth up to €150, charged per tariff classification rather than per parcel, running until 1 July 2028, with the seller or importer responsible for declaring and paying it. The Commission’s worked example is the clearest statement of the per-classification rule anywhere: five T-shirts attract €3 because they share one classification, while three T-shirts and a watch attract €6 because that is two.

EU card fees are lower than US ones, and the sources disagree about how much: 2026 fee breakdowns report 2% + €0.25 as a general EU Basic rate, with country figures reported separately at 2.1% + €0.30 for Germany and 1.5% + €0.25 for France, while Shopify’s own pricing-overview page says only that rates vary by plan and region and prints no numbers at all. Take yours from your own admin rather than from anyone’s table, including ours.

Now put it through the formula. A €9 landed cost, EU card fees at 2% + €0.25, a 10% return rate, a 15% target net margin, so a divisor of 0.73. The ad cost is a placeholder of €30, chosen as a round working number rather than converted from the dollar benchmarks, because we are not going to invent an exchange rate and call it research. Substitute your own.

Sourcing routeDuty per orderPrice required
EU stock, nothing crosses a border€0€53.77
Shipped from outside the EU, one classification€3€57.88
Mixed bundle, two classifications€6€61.99

Each €3 of duty raises the price you need by €4.11, not by €3. That is the part every duty explainer misses, including ours until now. A fixed cost added to an order has to carry its own share of the card percentage, the return risk and the margin you wanted, so it arrives at the price line multiplied by 1 divided by 0.73, which is about 1.37. The same multiplier applies to anything else fixed you add: a €2 insert card is €2.74 of price, a €4 shipping upgrade is €5.48.

It also puts a number on the standard advice to raise average order value by bundling. Bundle three units of the same product and the duty stays at €3. Bundle a gadget with an unrelated accessory and you have bought €4.11 of extra price requirement along with the higher basket. Check your real figure with the EU landed cost calculator, read the duty explainer for the mechanics, and note that EU stock removes the line entirely while also cutting the delivery times that drive returns.

Your sticker price is not your price: VAT

Everything so far is the money you keep. In the EU, the number on the product page has to be the number the customer pays. We read the Price Indication Directive 98/6/EC directly: Article 2(a) defines the selling price as “the final price for a unit of the product, including VAT and all other taxes”, and Article 4 requires it to be unambiguous, easily identifiable and clearly legible. Adding tax at checkout, which is normal in the US, is not an option for EU consumers.

So the €57.88 above becomes a sticker of about €70.56 at the EU average standard rate of 21.9%, €67.72 in Luxembourg at 17%, €72.35 in Denmark at 25% and €73.51 in Hungary at 27%. Those rates come from the Tax Foundation’s 2026 VAT table, which also lists Finland at 25.5% and Malta at 18%.

Here is the consequence nobody warns beginners about. Most stores show one price across the EU, which means the VAT is coming out of your margin at a different rate in every country. On a single €69.99 sticker you net €59.82 from a Luxembourg buyer and €55.11 from a Hungarian one, a €4.71 swing on identical orders. That is larger than the €4.11 the €3 duty costs you, and it arrives with no announcement, no news coverage and no line in your dashboard. Price against your realistic mix of destination countries, not against your cheapest one. This is general information rather than tax advice, and the Europe guide covers registration and OSS.

The compare-at price is regulated, and most dropshipping stores break the rule on day one

Every Shopify theme has a “compare at price” field, and the standard dropshipping move is to fill it with an invented higher number so the real price looks like a discount. In the EU that is illegal.

The Omnibus Directive (EU) 2019/2161 amended the Price Indication Directive so that any announced price reduction must state a prior price equal to the lowest price you actually applied in at least the 30 days before the reduction. Shopify’s own merchant documentation states it plainly: when you announce a price reduction to customers in the EEA you must display the lowest price you previously charged in the last 30 days or more, the rule covers previously reduced promotional prices inside that window, and some member states require a longer period. The Commission’s illustrative case is a “40% off” claim where the lowest price in the previous 30 days was €100: the discount must be calculated from €100 even if the last selling price was €160.

Sourcing honesty on this one. We read the VAT-inclusive definition in the directive text itself, but the consolidated version carrying the inserted Article 6a would not load in this session, so the 30-day rule above comes from Shopify’s merchant guidance and legal commentary rather than the legal text. The date, the substance and the 30-day period are consistent across every source we checked. Confirm the implementation detail for your own country before you run a campaign on it.

The practical effect on pricing is simple and slightly liberating: a new store has no 30-day price history, so it has no honest discount to announce. Launch at the price the arithmetic gave you, leave the compare-at field empty, and earn the right to run a real sale after a month of trading.

FAQ

How do I price a dropshipping product?

Work backwards from what a customer costs to acquire rather than forwards from your product cost. Price = (landed cost + duty + fixed card fee + target ad cost per sale) divided by (1 minus card percentage minus return rate minus target net margin). With a $12 landed cost, US card fees of 2.9% + $0.30, a 10% return rate and a 15% target net margin, the divisor is 0.721. At the $38.19 median Meta cost per acquisition reported in Triple Whale's 2025 dataset of nearly 35,000 brands, that formula returns a price of about $70, before the US import duty that every order shipped from abroad now owes.

Is the 3x markup rule still good enough in 2026?

As a floor it is fine, as a method it points the wrong way. Markup rules set your price from your cost, but the largest number in the calculation is the cost of finding a customer, and that barely changes when your price does. A $25 product priced at 3x can afford about $5.73 to buy a customer, against a median Meta cost per acquisition of $38.19. The rule can produce a perfectly obedient price that no advertising budget can support.

How much does US import duty add to my price?

More than its rate. Duty-free entry for parcels under $800 ended for goods from China on 2 May 2025 and for every country on 29 August 2025, so every order shipped to your customer from abroad owes duty on the item's value. As of October 2026 a China-made product pays its base duty, its Section 301 list (7.5% on List 4A, where most consumer goods are, 25% on Lists 1 to 3) and a 12.5% layer added on 24 July 2026, so a List 4A product with no base duty pays 20%. Because the duty also has to carry its share of your card percentage, return risk and target margin, each dollar of it raises the price you need by about $1.39. A product shipped from US stock has no duty line per order, because the importer already paid it.

How much does the EU €3 duty add to my price?

About €4.11, not €3. Since 1 July 2026 the EU charges a flat €3 duty per tariff classification on parcels arriving from outside the EU on goods up to €150, running until 1 July 2028 per the European Commission. Because that fixed cost also has to carry its share of your card percentage, return risk and target margin, it arrives at the price line multiplied by roughly 1.37. A mixed bundle spanning two classifications is charged €6 and costs about €8.22 of price. Sourcing from EU stock removes the line entirely.

Should my displayed price include VAT?

In the US there is no VAT, and prices are normally shown before sales tax, which is added at checkout. For consumers in the EU, yes. Article 2(a) of the Price Indication Directive 98/6/EC defines the selling price as the final price for a unit of the product including VAT and all other taxes, and Article 4 requires it to be unambiguous and clearly legible. Adding tax at checkout is not an option. Because standard rates run from 17% in Luxembourg to 27% in Hungary, a single EU-wide sticker means your net revenue varies by country: on a €69.99 price you keep €59.82 from a Luxembourg buyer and €55.11 from a Hungarian one. This is general information, not tax advice.

Can I show a crossed-out compare-at price on a new store?

Not honestly, and not legally in the EEA. The Omnibus Directive (EU) 2019/2161 requires any announced price reduction to state a prior price equal to the lowest price you actually applied in at least the 30 days before the reduction, and Shopify's own merchant documentation repeats this for EEA customers, noting that some member states require a longer period. A brand new store has no 30-day price history, so it has no discount to announce. Launch at your calculated price, leave the compare-at field empty, and run a real promotion after a month of trading.

Do prices ending in 9 actually sell better?

The direction has real experimental support and the popular percentages do not. Anderson and Simester published three field experiments in Quantitative Marketing and Economics in 2003 in which a $9 price ending increased demand in all three, with a stronger effect on items new to the retailer than on items sold in previous years, which is the relevant case for a dropshipper because your product is new to every visitor. The widely quoted 24% lift is not from that study and cannot be traced to an inspectable source, and reviews of the wider literature report a mix of positive, null and negative results.

Should shipping be free or charged separately?

Build it into the price and show one number. Baymard Institute's meta-analysis of 50 studies puts documented cart abandonment at 70.22%, and excluding people who were only browsing, the top reason is extra costs being too high at 40%, ahead of slow delivery at 20%. Splitting the price into product plus shipping gives the customer two separate chances to refuse. The formula is indifferent to how you split it because shipping to the customer is already inside the landed cost line.

Done with this module?Next: Making the ad, 9 min
Back to the course