A customer bought from you last year. This half, they tried Shein or Temu for the first time. The quick read is that you lost them to a price you cannot match. The latest CACE study says something different: for most of them, that purchase abroad was a one-off, not a habit change.
What the CACE study actually measures
The Mid Term 2026 study, run by Kantar for the Argentine Chamber of Electronic Commerce (CACE), covers the first half of the year: $19.53 trillion pesos in revenue (up 28% year over year), 181.5 million orders (up 21%), and an average ticket of $107,597 pesos. Within that, 18% of respondents made their first international purchase during the half. But among those who buy abroad, 64% do it occasionally, and only 10% do it habitually.
That 64% is the number that matters to a local ecommerce, more than the headline that Shein and Temu are growing. Most people who tried an international platform did not turn that trial into a routine. They did not stop buying from you to buy from someone else for good: they tried it once, usually for the price, and the question that matters is who wins the next order.
Why “occasional” is not the same as “lost”
A one-off purchase abroad does not erase the history a local business already holds on that customer: what they bought before, how often they come back, which size or variant they prefer. That history is exactly what a customer data platform exists to hold in one place instead of scattered across systems that do not talk to each other, and it is what an international platform will never have on them. Shein and Temu compete on price for a single transaction; a local business competes with something those platforms cannot offer: a relationship with history.
What to do with the data instead of competing on price
- Find the customers who bought once and did not come back within the window when they would normally repeat, not the whole base: that is where the one-off trial lives, mixed in with ordinary churn.
- Use that purchase’s history (category, variant, lifecycle stage) for the next message, instead of a generic discount sent to the whole list.
- Pick the timing, not just the message: the window when that customer is normally due for another purchase, the same principle that separates a real second purchase from one that never comes.
- Measure that cohort separately from the overall repeat-purchase rate: if it recovers differently from the rest of the base, that confirms it was a one-off trial, not a habit change.
If your answer to Shein or Temu is cutting price across the whole category, you are fighting the fight that favors you least. The one you can actually win is knowing, better than any international platform, when and what that customer is due to buy next.