The Ecomm Analyst

The first time I looked at a multi-market store’s attribution data, the blended ROAS was wrong by roughly nine percent and nobody could tell me why. The answer was boring. The store sold in six currencies, the ad accounts were denominated in three, and the attribution tool was converting all of it to USD at a rate that had nothing to do with what actually settled in the bank.

This is the part of international selling nobody warns you about. Your storefront shows a customer in Germany a price in euros. Depending on your payment setup, you may settle in euros or you may settle in your home currency at a rate the processor picked. Your ad platform reports spend in whatever currency that ad account was created in. Your attribution tool then has to reconcile all three, and how it does that determines whether your ROAS is a real number or a rounding error with a decimal point.

1. RedTrack

RedTrack treats this as a first-class feature rather than an enterprise upsell. Multi-currency support is included on both the Brand and Agency e-commerce plans, revenue syncs in real time across every connected market, and reports display in the currency you choose. No manual conversion step, no reconciliation lag between what the store earned and what the dashboard shows. Brand is published at $99 per month, or $83 billed annually, tiering by annual tracked revenue from the up to $250K band upward, and includes three stores with additional stores at $29 each per month. If you are running separate country stores rather than one store with multiple markets, that store allowance is the number to check first.

2. Wicked Reports

Wicked Reports gates currency conversion behind its Scale plan, which at the entry revenue band is $699 per month against $499 for Measure. Scale also adds international time-zone data loading, and that second item is quietly the more valuable one. If your tool loads data on US Eastern time and your biggest market is in Australia, a meaningful share of your orders land on the wrong calendar day, and every day-over-day comparison you make is slightly fictional. Wicked is transparent that complex multi-brand setups need a conversation with their team. The pricing is high enough that I would want the LTV and cohort reporting to be the reason I am there, with currency handling as the thing that makes it usable rather than the thing I am paying for.

3. ThoughtMetric

ThoughtMetric, which sponsors this blog, takes a simpler approach and I want to describe it accurately rather than generously. You set a store currency in settings, it defaults to USD, and ad spend and revenue report in that single currency. That is a consolidated view, not a per-market currency view. For a brand selling internationally out of one store and one set of books, that is usually what you actually want, and it avoids the trap of reading six dashboards that each tell a different story. If you need to compare market-level performance in local currency side by side, this is not the model you want. Pricing starts at $99 per month for 50,000 pageviews with every feature at every tier and a two-week trial, and the back-end order connections cover refunds and cancellations, which matters more in international selling where return rates run higher.

4. GA4

GA4 handles multiple currencies, and the way it handles them is worth understanding before you trust the output. You set one property currency. Google converts every non-USD transaction to USD at processing time using the previous day’s exchange rate, then converts back to your property currency when the report is generated. Google’s own documentation acknowledges this produces small discrepancies even when your property currency already matches your local currency. In practice I see roughly one to two percent drift against internal numbers, which is small until you are arguing about whether a channel cleared break-even. It is free and it is a fine cross-check. It is not the source of truth for a multi-market P and L.

What to test before you sign anything

Pull one week of orders from a single non-home market and reconcile them by hand against the tool. Three numbers have to line up. The order total in the currency the customer paid, the amount that settled, and what the tool reports. If the tool converts at a daily rate and you settle at a monthly average, you will never get exact agreement, and that is fine as long as you know the size of the gap and it stays stable.

Then check the time zone setting on every platform in the stack and make them match. It is the least interesting fix available and it resolves more international reporting disputes than anything clever.

The underlying problem here is the same one I wrote about when building one revenue number you can run the business on. Currency just adds a layer where two systems can disagree while both being technically correct.