Foreign shoppers in 58 international markets may be shown your checkout in USD, not their own currency, and with no local way to pay.
A currency leak is simply this: a shopper who can’t see or pay in their own money hesitates, second-guesses the price, and often abandons the cart. Those are sales you never see leave.
Neutral scan | findings are candidates to verify from each marketRecoverable per year | modeled
modeled opportunity | verify per market | +4 to 13% international conversion | on the 30% international share you gave us.
Markets scanned: United States, United Kingdom, France, Italy, Spain, Canada, Australia, Singapore, Sweden, New Zealand, Hong Kong, Switzerland, Germany, Netherlands, Belgium, Poland, Brazil, Mexico, Argentina, Colombia, India, China, Japan, South Korea, Indonesia, Thailand, South Africa, Nigeria, Saudi Arabia, United Arab Emirates, Norway, Denmark, Finland, Portugal, Austria, Ireland, Czechia, Greece, Türkiye, Malaysia, Philippines, Vietnam, Taiwan, Egypt, Kenya, Chile, Peru, Romania, Hungary, Slovakia, Bulgaria, Croatia, Lithuania, Slovenia, Latvia, Estonia, Luxembourg, Cyprus, Malta.
The proof | what you see vs what the world seescandidate | verify
You only ever check out from your own country, in your own currency. Your foreign shoppers don't.
The fix | start here
Turn on local presentment currencies: each market sees and pays in its own.
What this scan could not do
It read your checkout from one location, so every finding above is a candidate. We can reopen it from inside each of the 59 markets through a residential connection there, and send you the screenshot of the page your shopper actually gets. See what that costs →
Your assumed international revenue, split across the scanned markets by published global e-commerce weight per market. Swap in your real country mix and the map redraws; the total stays the same.
| 01 | China to verify | no local payment method, no local currency | ~$41,124/yr $19,777 to $63,863 |
| 02 | Japan to verify | no local payment method, no local currency | ~$4,112/yr $1,978 to $6,386 |
| 03 | Germany to verify | no local payment method, no local currency | ~$3,290/yr $1,582 to $5,109 |
| 04 | South Korea to verify | no local payment method, no local currency | ~$3,290/yr $1,582 to $5,109 |
| 05 | United Kingdom to verify | no local currency | ~$3,076/yr $1,479 to $4,777 |
| 06 | India to verify | no local payment method, no local currency | ~$2,467/yr $1,186 to $3,832 |
| 07 | Brazil to verify | no local payment method, no local currency | ~$1,645/yr $791 to $2,555 |
| 08 | France to verify | no local currency | ~$1,538/yr $740 to $2,388 |
| 09 | Canada to verify | no local currency | ~$1,025/yr $493 to $1,592 |
| 10 | Netherlands to verify | no local payment method, no local currency | ~$822/yr $395 to $1,277 |
+ 48 more markets affected.
Rows show the mid of each market's modeled band; the per-market lows and highs sum exactly to the headline band's endpoints, so a single market's mid can exceed the conservative floor.
Split driver: published e-com weights. Replace with your analytics (Shopify Admin → Analytics → Sales by billing country). If you don't sell to a listed market, strike it: its dollars reallocate to the others, and the total is unchanged.
The percentage is the model's output. Your money is just that % applied to your own revenue. Find your row:
At your ~$250,000/mo, that's $37,356 to $120,631 recoverable per year.
| Monthly online revenue | Recoverable / year |
|---|---|
| $100,000/mo | $15,120 to $48,240 |
| $250,000/mo | $37,800 to $120,600 |
| $500,000/mo | $75,600 to $241,200 |
| $1,000,000/mo | $151,200 to $482,400 |
| $2,000,000/mo | $302,400 to $964,800 |
| Intl share | Recoverable / year |
|---|---|
| 15% | $18,678 to $60,316 |
| 30% (assumed) | $37,356 to $120,631 |
| 45% | $56,034 to $180,946 |
The estimate scales linearly with your international share; the assumed 30% row matches the figures above.
Assumes ~30% of revenue is international | scale linearly to your real figure. Scenario rows are the % band applied to round revenue figures; your exact band is above.
Ranked by modeled recovery (candidates | verify per market), with where the fix lives and how fast it pays back. Most are a same-week change.
The signals point to a modeled ~$37,356/year on the table, to be verified from each market, which is exactly what the audit does. 89 CHF pays for itself in ~1 day, even at that modeled floor.
The audit names every leak, gives you numbered fixes for your platform, and a step-by-step per-market verification protocol you can run in minutes.
Each block is a self-contained work order: paste one per ticket into Jira / Linear. The candidate / modeled labels carry over from the report; don't drop them when you forward.
We could not reach these markets from inside the country this scan, so every finding in them is a candidate. Run the checks below and record what the page actually shows. A re-scan applies the same criteria.
We scanned https://aurora-outfitters.example/cart's checkout across 59 markets on sample data; geo verification: none | candidates. 33 leaks; checkout health 65/100. A modeled $37,356 to $120,631/yr opportunity, to verify per market. Start with: the top fix in the plan, unlocked in the full report.
Currency leaks come back on their own:
Your PSP can silently re-enable DCC after an update.
Opening a new market re-introduces a forced currency.
A theme swap can silently drop your currency selector: Markets stays on, shoppers stop seeing it.
FX rates drift every single day.
That's why merchants keep us on Checkout Watch (39 CHF/mo): we re-scan every market monthly and alert you the moment a leak reappears. One month of the modeled floor left open is ~$3,113 (modeled). Weigh that against the subscription.
Why a human, not just a tool: the scan finds candidates; we open your checkout from the real market, verify each leak, and hand you the proof, so you act on facts, not false alarms.
Names every leak, numbered fixes for your platform, and a step-by-step per-market verification protocol you can run in minutes. Pays for itself in ~1 days, even at the modeled floor (to verify).
Send me the proof →Every figure here is traceable to a disclosed assumption: no black box, no AI. Change the inputs to your real numbers and the estimate moves with them.
| Leak type | Published stat → haircut | Band used |
|---|---|---|
| hidden FX markup (DCC) | Published DCC markups run 3 to 7% over the interbank rate; only a share of shoppers notice and balk, so the realised band is haircut well below the markup itself. The haircut is our assumption. Source: Mastercard / Visa DCC markup data. | +3 to 10% |
| stale FX rate | No study cleanly isolates stale-FX drag, so we price it well below the forced-currency band: the padding is smaller than a wholly foreign price. The band is our assumption. Source: our assumption | no single published stat. | +1 to 4% |
| no local currency | 33 to 49% of cross-border shoppers say they abandon without local currency; we haircut that stated intent hard (roughly 3×) to a realised uplift band. The haircut is our assumption. Source: PYMNTS / Passport, 2025/26. | +5 to 15% |
| surprise duties at delivery | Surprise costs at delivery are a repeatedly stated cross-border abandonment driver; without one clean published number we price a small, conservative band | our assumption. Source: our assumption | no single published stat. | +2 to 7% |
| tax-exclusive prices (VAT surprise) | Unexpected extra costs at checkout are the top stated abandonment driver (Baymard-class studies), but they bundle shipping with taxes, so we price the tax-display share of it as a small conservative band | our assumption. Source: our assumption | no single published stat. | +1.5 to 5% |
| no local payment method | PPRO reports a material conversion lift when a market's dominant local method is offered; we haircut that to a realised band because card-comfortable shoppers still convert. The haircut is our assumption. Source: PPRO local-payment-method studies. | +3 to 12% |
Bands are realised-recovery assumptions applied to the affected markets' international revenue share. Where a haircut is our judgment, the sentence says so.
Deterministic | parsing + interbank-rate arithmetic | no AI | fully reproducible
Scan sample data | engine v1.1.0 | FX: interbank mid-market at scan time
This report is an automated, deterministic estimate for informational purposes only. It is not financial, legal, or tax advice, and not a guarantee of results. Figures depend on the disclosed assumptions; verify against your own analytics before acting.