Most international expansion fails for one reason. The brand brings a blended view of its numbers into a market that needs its own. Rhetica is a B2B international-expansion and DTC growth consultancy that builds and operates new-market revenue.
The pattern is the same across categories. A healthy global average hides at least one market that loses money on every order.
This is the hub for our finance and unit-economics work. It explains the four numbers that decide whether a market is worth entering. Then it links down to the deep-dive for each one. Read this first, then go deep where you need to.
I have watched brands ship into a new country on the strength of a clean dashboard, then spend six months learning the dashboard was lying. The fix is not more spend. It is four numbers, read one market at a time.
What unit economics decide whether to expand into a new market?
> Four numbers decide whether a market is worth entering: true CAC (the fully loaded cost to acquire a customer there), contribution margin by country, payback period from that market’s own P&L, and the AOV-to-CAC ratio. Rhetica’s rule is to model all four per market, never blended, because a healthy global average routinely hides markets that lose money.
The reason this matters is mechanical. A blended number is an average. An average can be carried by one strong market while a second one bleeds. Your dashboard looks fine, but your bank balance does not.
Modeling per market is the only way to see the market that is dragging. You want to see it before it has spent six months of budget proving the point. By then the cash is gone and the lesson is expensive.
The four numbers build on each other. True CAC feeds contribution margin. Contribution margin feeds payback.
The AOV-to-CAC ratio is the fast screen you run before any of the heavier modeling. Here is each one.
True CAC: the fully loaded cost to acquire a customer there
Domestic CAC is one line: ad spend divided by new customers. True CAC in a new market is that line plus everything the new market adds. Creative has to be localized, not just translated. Payment methods convert at different rates and carry different fees.
Return rates vary by category and by country. Cross-border fraud and chargebacks take a cut. Local-language support is often a real per-customer cost in year one. Each of these is small on its own, but they stack fast.
Take a DTC brand whose US CAC is, say, roughly $40 (illustrative). It enters a new European market and runs the same playbook. Localized creative adds a few dollars per customer once spread over a year of volume. A local payment method converts worse and costs more, adding a few more.
Returns in the category run higher there. Stack those lines and the true cost to acquire is no longer $40. It can land closer to $70 or $80 (illustrative). The brand never sees this if it reports one blended CAC, because the cheap US customers pull the expensive new-market ones back down.
When we model a market for a client, we build these cost lines before the first dollar of spend. We carry them per geo, not as one number.
The figure you get is the only honest input to everything downstream. Get this wrong and the next three numbers are wrong too.
Contribution margin by country: what each order actually keeps
Contribution margin is what is left from an order after the costs that scale with that order. Those costs are goods, shipping, duties, payment fees, returns, and the per-order share of fulfillment. It is not gross margin. Gross margin flatters you because it ignores the costs that a new country quietly inflates.
Cross-border changes almost every line. Duties and import costs raise the landed cost. Shipping a physical good to a new region costs more and takes longer. Local taxes and payment rails take their share.
A product that keeps, say, around 60 percent contribution at home (illustrative) can keep far less abroad once duties and freight are in. Sometimes the same SKU at the same price is a strong product in one country and a thin one next door. The only difference is what it costs to land and deliver there.
Model this per country and per SKU, not as a regional blend. The brand that averages a strong home market with a duty-heavy new one will see a comfortable number. That number does not exist in either place. It is the math hiding a problem, not solving it.
Payback by market: read it from that market’s own P&L
Payback period is how long it takes the contribution margin from a customer to repay the true CAC of acquiring them. The mistake is reading it from the blended book. You have to read it from the market’s own P&L. Use that market’s true CAC and that market’s contribution margin.
Work an illustrative example. Suppose true CAC in the new market is about $75 and the contribution margin per order is roughly $25 (illustrative). On a single purchase, that customer has not paid back. They are underwater until repeat purchases close the gap.
Repeat behavior in a new market is an assumption until you have data. It is not a fact you can borrow from home. A market that paid back in three months at home can stretch past a year abroad. That happens once true CAC is higher and contribution per order is thinner.
Payback is where working capital bites. A longer payback means more cash tied up in each cohort before it returns. Two markets can show the same lifetime value and behave very differently on cash. The per-market P&L is the only place that shows up before it hurts.
AOV-to-CAC ratio: the fast screen that decides entry
The AOV-to-CAC ratio is the quick test you run first. Take average order value in the new market and divide it by true CAC there. It tells you, on a single order, whether the economics are even close. You run it before you commit to the heavier modeling.
A ratio well above one on the first order means the market can plausibly work. A ratio near or below one means the first order does not cover the cost to acquire. So the whole thesis rests on repeat purchase. That is the riskiest assumption in a market you have never operated in.
Say AOV in the new market is around $90 and true CAC is roughly $75 (illustrative). That is a ratio near 1.2. It is thin enough that the market only works if repeat behavior holds up. Compare that to a home market where AOV is $90 against a $40 CAC, a ratio over two, and the gap in risk is obvious.
This ratio does not replace the other three numbers. It screens out the markets that should never have been on the list. That way you spend modeling effort on the ones with a real chance to pay.
The five deep-dives
Each number above has its own deep-dive. Start with whichever one is your current bottleneck.
- how to calculate true CAC in a new market. The full cost lines that turn a domestic CAC formula into an honest per-geo number.
- contribution margin by country. How duties, freight, payment rails, and returns reshape margin SKU by SKU and country by country.
- building a per-market P&L and payback. How to read payback from a market’s own P&L and why working capital is the number behind the number.
- the AOV-to-CAC math that decides entry. The fast ratio that screens markets in or out before you model anything else.
- the expansion thesis your CFO will fund. How to assemble the four numbers into a per-market case that survives a finance review.
The verdict
Run all four numbers per market and the decision usually makes itself. The market either clears true CAC, holds contribution margin, pays back inside a window you can fund, and shows a workable AOV-to-CAC ratio, or it does not. There is not much room in between.
Skip the per-market work and you are betting the blended average is telling the truth. In my experience operating these launches, it almost never is. Build the four numbers first, then decide. If you want help building them for your markets, that is the work Rhetica does.