Year one of cross-border ecommerce is forgiving. The launch attracts attention. The team is energized. Investor decks include pleasant slope charts. Revenue grows faster than cost grows, because the cost is largely upfront and the revenue is recurring. Almost any reasonable team can make year one look like a success.
Year two is where the structural mistakes start charging interest. The novelty premium fades. Customer acquisition cost catches up with the spend that built the launch. The fulfillment partner you picked because they were available shows the operational ceiling you didn't notice. The marketplace you chose for ease begins to commoditize your margin. The localization shortcuts you took in week three become the friction your customer support team manages indefinitely.
Three failure modes we see most often.
1. Channel concentration that masquerades as channel strength.
Many year-one launches are 80% Amazon JP or 80% Rakuten. That can be the right call at the start — marketplaces give you reviews, distribution, and search demand without paid media. But by month fifteen, the brand is a tenant, not a landlord. The marketplace owns the customer relationship, the data, and increasingly the pricing. Brands that don't begin building DTC and retail diversification in year one find themselves negotiating margin with a counterparty that has no incentive to give.
2. Operations stitched together at launch and never refactored.
The 3PL chosen for capacity rather than fit. The integration done in Zapier rather than properly. The customer service handled by a freelancer in three timezones. None of these are mistakes at launch — they are pragmatic — but they are debts. Year two is where the debts come due in the form of stockouts, missed shipments, and customer churn that doesn't show up cleanly in the dashboards.
3. Localization that fades when the launch attention does.
At launch, your hero pages are localized, your packaging is right, your social posts are written by a native copywriter. Six months in, those things are being maintained by the same team that maintains your home market — and the cracks show. Old promotion templates pushed to a market where they don't work. Email subject lines translated by an AI tool that doesn't know your category. The brand starts to read foreign again, just at the moment it should be feeling native.
What good year-two thinking looks like.
The brands that turn year two into a flywheel make three structural moves.
- Channel diversification with intent. By month fifteen, marketplace concentration has dropped from 80% to roughly 50%, with DTC and retail picking up share. Not because marketplaces are bad, but because the brand is no longer a captive of any one of them.
- Operational refactoring before scale forces the issue. Migrating from launch-grade tooling to the platform that will carry the next five years of revenue. Not glamorous work, but the work that lets year three be uneventful.
- Local team or local partner with continuity. Either an in-market hire who owns the market roadmap, or a partner agency on a continuity contract. The model that fails most reliably is rotating freelancers and project agencies handed off every two quarters.
"The structural mistakes of a year-one launch don't show in year one. That's the trap. Year two is where the math starts being honest with you."
The honest pre-mortem.
Before any cross-border launch we run with a client, we do a pre-mortem on year two. We write the story of how the launch fails in eighteen months and reverse-engineer what we'd have to get wrong to make that story true. Then we structurally avoid those things. It is the cheapest insurance policy in cross-border ecommerce. It is also one almost no one writes.
Year two is not a marketing problem. It is the year your operating model is graded. Treat the first year accordingly.
Deebo is an international expansion and cross-border ecommerce agency, headquartered in Tokyo, working with foreign brands across Japan, APAC, and 40+ markets worldwide.