Before going into this imperative of increasing service levels, what were the levers to push the $2.50 towards the $2.20? What was still available to pursue?

Basically, there was a path to increase drivers by removing drivers from 3PLs for first mile. There were around 20% of drivers from 3PLs, and when we removed 3PLs, you would save around 50% of costs. Shopee's pricing structure for its own drivers was to pay by parcel, whereas with 3PLs you had to pay by route and per truck usage per day. That would give you $0.05 at most. The other big part was in line haul. Only 20% of line haul was Shopee contracting directly with the driver, and only for the shortest routes. For line haul, there was still around $0.10-$0.15 in the long run to reduce. Combined with the first mile savings, that is around $0.20, so that gets you to $2.30.

This is more on the strategic side. How was the cost of delivery seen internally in relation to the average selling price of the merchandise at Shopee? What stands out to me is that Shopee is offering free shipping for packages of 10, 15, 20 reais, and the delivery cost is 10 to 15 reais, so it is very uneconomic to serve this segment of the market. How did these discussions take place internally?

When I joined Shopee, the average ticket was $6-7. When I left in early 2025, it was $13 per order. The free shipping threshold was 20 BRL for Brazil and 10 BRL for São Paulo. People actually spend more than the threshold. If you consider the 20% take rate on $13, that's $2.60, plus fixed fees of about 7 cents per dollar per item sold on top of the 20%. When you sum that up, there is profitability, though it's small—less than 10 cents per order. Logistics was by far the biggest cost. Other costs like credit card transactions are negligible.