A Big Mac costs $6.22 in the United States and $2.42 in Taiwan — same sandwich, same month, the July 2026 edition of the index this site mirrors. Nothing about the dollar changed on the flight over. What changed is everything the burger is assembled from once it lands: the rent, the wage of the person building it, the beef.
I bring that up because I sell a paid app, and Apple prices it in roughly 175 storefronts by converting one US number at something close to the market exchange rate. When I finally divided each of those local prices by the local Big Mac, the same download ran from 0.40 burgers in Turkey to 1.70 in Vietnam — a 4.25x spread I never chose. The full table, and how I ended up staring at it, is the first piece in this series.
This one isn’t about apps.
The arithmetic doesn’t know what it’s dividing. An ebook, a Notion template, a seat-based SaaS plan, a $200 cohort course — anything with a marginal cost near zero and customers in more than one economy carries the same problem, usually for the same boring reason mine did. The default was already ticked.
Three things are worth pulling apart: why an exchange rate is the wrong instrument, why a purchasing power parity discount is a pricing decision rather than a favour, and where the whole model quits on you. Everything below is one app, one month — May 2026 sales, July 2026 burger prices. Sales are indexed to the United States at 100, because the absolute unit counts aren’t mine alone to publish. Read it as a case, not a study.
Exchange rates versus purchasing power
An exchange rate is a price set by people who trade money for a living. Capital flows, rate differentials, central-bank policy, somebody’s Tuesday hedge. What a designer in Istanbul earns in a month never enters that market — not directly, not indirectly. It isn’t an input.
So when a storefront converts $5.99 into lira, it’s answering a question nobody asked — what would this cost if a currency desk bought it? The question you meant to ask is what it costs the person actually holding the lira.
The meaning of PPP pricing fits in one line — you set each local price so it costs roughly the same slice of local life as the home price does at home, instead of the same number of converted dollars. The Big Mac is the cheapest published way to see the gap between those two things. It’s the same physical object on both counters, and it’s built almost entirely out of inputs that never cross a border.
July 2026 edition, straight off the CSV. United States $6.22. Switzerland $9.04. China $3.91. Taiwan $2.42. India $2.45 — and India’s is chicken, which is one of about seven reasons this instrument leaks.
Now the test that catches people. A Turkish Big Mac costs $6.91 — more than an American one. Turkey isn’t a high-income country by any measure you’d care to pick, and yet the burger there costs more than it does in the United States, while my app converted down to $2.74. The lira did that. I didn’t.
That’s the first principle underneath every regional pricing strategy anyone has written down — money isn’t the same size everywhere, and the market rate isn’t the instrument that tells you how big it is where your customer lives. You can run any currency pair through the purchasing power calculator and watch the market rate and the burger rate disagree.
The parity pricing gap
Here’s the whole check. It takes about ten minutes for any product with a public price list.
Write down your reference price — the one you actually reasoned about. Mine is $5.99, and I reasoned about it for maybe ninety seconds.
Divide every local price by that country’s Big Mac. One division per market. The unit that comes out is “how many burgers does this cost”, which sounds like a joke until you sort the column.
Read the ends, not the middle. The middle is reassuring and useless. On my list the US, Thailand, Germany and Singapore all landed between 0.94 and 0.97 — four countries with nothing in common, agreeing. That isn’t care on my part. Those are simply the places where the market rate sits close to purchasing power. Where the two disagree, my price went wherever the currency took it.
Then decide which spread you meant to have. Not zero, necessarily. Just one you chose.
An extract from my own column, July 2026 burger prices against public App Store prices:
| Market | App price | ≈ USD | Big Mac | App in Big Macs |
|---|---|---|---|---|
| Turkey | ₺129 | $2.74 | $6.91 | 0.40 |
| China | ¥21.80 | $3.22 | $3.91 | 0.82 |
| United States | $5.99 | $5.99 | $6.22 | 0.96 |
| Britain | £5.99 | $8.07 | $7.40 | 1.09 |
| Japan | ¥790 | $4.87 | $3.08 | 1.58 |
| Vietnam | ₫129,000 | $4.91 | $2.89 | 1.70 |
Twenty markets are in the full table in part one. The shape doesn’t change — a Vietnamese buyer pays 4.25 times what a Turkish one does, measured in the one product that’s physically identical in both countries.
I built an App Store pricing calculator to run this across every Apple storefront at once. By hand in a spreadsheet, it’s the sort of chore you do exactly once.
A discount that isn’t charity
Is a purchasing power parity discount a favour to poorer countries? No. It’s price discrimination, which is the oldest respectable idea in pricing and has nothing to do with generosity.
Student tickets, matinee seats, the cheap paperback edition that comes out a year later — all the same move. You’re not donating a unit. You’re selling one you otherwise wouldn’t sell, at a price above your marginal cost, to a buyer sitting on a different demand curve. For a digital product that marginal cost is close to zero, which is why the floor can go so low without the arithmetic breaking.
My clearest illustration is China. The price there is ¥21.80, about $3.22 — roughly 54% of the US price. Indexed to the United States at 100, May 2026 net sales came in at China 159. Second place was Taiwan at 19, then Japan 14, Mexico 9, Britain 8, Saudi Arabia 8.
So the storefront paying about half the American price outsold America by 1.59x — and outsold every other English-speaking market on my list by more than an order of magnitude.
I want to be careful about what that does and doesn’t show. It’s one app in one month, it’s a drawing toy rather than a tool anybody needs, and I ran no experiment — the lower price is tangled up with a hundred other things about that market. What I can say is narrow — the cheapest major storefront on my list was also the biggest one, and nothing in my numbers supports the idea that a PPP discount is money thrown away.
And it runs the other way too, which is the half people forget. Saudi Arabia sits at 1.32 burgers — well above the American 0.96 — indexed 8, with a refund rate of 0.0%. Parity pricing isn’t a synonym for cheaper. In about half the world it means charging more than the converted US price, and the market where I charge the most in real terms is the one that asked for nothing back.
The refund column tempts you to say more than it can carry. Britain 20.0%, Germany 16.7%, Thailand 14.3%, China 6.9%, the United States 0.0%. Britain’s 20.0% is a couple of refunds against a denominator small enough that I could move it two points by sneezing. That’s not evidence. It’s an anecdote with a decimal point — and I’d rather say so than let a bold percentage do work it can’t carry.
One more thing I got wrong, since I’m handing out advice. I spent years assuming Nordic buyers were the best customers on earth — high income, high standards, comfortable paying for quality. On that same 100-point scale — Norway 1, Finland 2, Sweden 0, Denmark 0. Switzerland, home of the $9.04 burger, came in under 1. So much for the IKEA theory.
Country specific pricing beyond apps
There’s an old phrase for what this is — 因地制宜, set your measures according to the ground you’re standing on. A price list is a claim about ground you’ve mostly never stood on.
How the method changes by product shape, in the order the mistakes get expensive:
Subscriptions. The error compounds monthly, and the failure mode is invisible — a market priced too high doesn’t churn loudly, it just never signs up. Nobody files a ticket about a purchase they didn’t make. The compensating advantage is real, though. You can move a price for new cohorts without a refund wave — which makes a regional pricing strategy for SaaS far easier to correct than a one-shot product’s.
Seat-based B2B. Your marginal cost isn’t zero — support, storage, payment processing, sometimes local tax handling. The PPP floor is your cost to serve, not the bottom of the burger table.
Ebooks, templates, one-off downloads. Often sold through a checkout with a single global price, which turns country specific pricing into a coupon problem rather than a price-list problem. That’s worse than it looks. A code is a sale, a sale creates a reference price, and the customer now knows the real number is the discounted one — part one goes into why repricing beats discounting.
Courses with a human in them. Every seat costs you hours. This is the case where a deep purchasing power parity discount can genuinely go below cost, and where I’d set a floor first and let PPP work above it.
Whatever you index against, be explicit about it. The Big Mac is public, monthly, and free — which is most of its appeal, and roughly all of its rigour. The Economist publishes a GDP-adjusted version of its own index precisely because the raw one is largely a wage indicator; the methodology page sets out where these prices come from and how far each tier can be trusted.
Where the model breaks
A PPP pricing model has three hard edges. The first one bites hardest.
Marginal cost. Everything above assumes the next unit is free. If serving a customer costs you money, the discount has a floor, and PPP will happily walk you straight through it — a market at half your reference price is one thing on a static file, and quite another on a product that bills you for storage and support every month it stays alive.
Arbitrage. A VPN and a foreign card, and your Turkish price is your global price. That’s real. I won’t wave it away. Two things bound it: the leak is limited to the share of buyers willing to do the work for a single-digit-dollar product, and the friction differs enormously by platform. A storefront tied to an account’s billing country is a different obstacle than your own checkout, where the whole job is a browser extension. If your gap is large and your friction is low, assume the gap gets found.
A public price list is a document your biggest customer reads. Consumer storefronts show each buyer their own number. A B2B pricing page shows everyone everything, and a wide visible gap becomes a procurement lever in every renewal conversation you have that year.
Then there’s the instrument itself. The Big Mac carries local labour, local rent, beef tariffs, McDonald’s own brand positioning — premium in some markets, price war in others — and in a fair chunk of the index the exchange rate is managed rather than floating. I’ve written the long version of that argument separately, in why PPP theory fails and the limits of the Big Mac index, and I’d rather you read it than take this article as a licence.
Which brings me to the line on my own list I like least. Switzerland — the most expensive burger in the July 2026 index at $9.04 — pays CHF 4.00 for my app, about $4.95. That’s less than an American pays in plain dollars, on a burger that costs 45% more. Call it half a Big Mac, against the American 0.96. I never decided that either.
I’ve been writing about how to catch exactly that, and it’s still sitting on my price list today.
FAQ
What does PPP pricing mean? Setting a separate price per country so that each price costs a roughly comparable amount of local purchasing power, rather than converting one price at the market exchange rate. Purchasing power parity is the underlying economic idea; the Big Mac index is one cheap, public illustration of it, and a leaky one — see the limits piece.
How large can the gap get if I never set country prices? On my own list it ran from 0.40 to 1.70 Big Macs — 4.25x — using public App Store prices and the July 2026 Big Mac index. That’s one product on one platform, not a general estimate. Your own spread depends on which countries you sell in and how your storefront rounds.
Does a PPP discount actually increase sales? I can’t answer that generally, and neither can anyone with one dataset. In my May 2026 numbers, China was priced at about 54% of the US price and indexed 159 against the United States at 100. There was no control group, no A/B test, and one month of one app — it’s consistent with the idea, and it doesn’t prove it.
Is PPP pricing the same as running a regional discount? Not in practice. A country price is a standing number a customer sees as the price; a discount is a temporary markdown against a reference price they now know exists. The mechanics of that distinction, and the evidence I could find on it, are in part one of this series.
Where do the Big Mac figures here come from? The Economist’s published Big Mac index, July 2026 edition, mirrored on this site with the raw CSV linked from every country page. The app prices are the public App Store prices, which anyone can check. Sales are indexed rather than absolute, and the refund percentages come off a small enough base that they should be read as anecdotes.