Shopee’s gross merchandise value grew 30.2% year on year in the first quarter of 2026, to US$37.3 billion. In the same quarter, adjusted EBITDA, Shopee parent Sea Limited’s own preferred profitability metric, fell 15.6%, from US$264.4 million to US$223.2 million. Both numbers appear in the same earnings release. The metric a platform puts at the top of its seller dashboard and the metric that actually decides whether the business made money did not just diverge that quarter. They moved in opposite directions.
That is not a scandal or a hidden problem. It is what GMV was always going to do once ad subsidies, fulfillment investment, and commission cuts get layered under a growth push. The issue is that a seller reading their own dashboard sees the same structural blind spot, at a smaller scale, every day.
What GMV actually counts
Gross merchandise value is the total dollar value of everything sold through a platform, before any fee, refund, discount, or cost is subtracted. It is a transaction-volume number, not a revenue number and nowhere close to a profit number. A US$50 order that nets the seller US$3 after commission, transaction fee, and ad spend counts exactly the same in GMV as a US$50 order that nets US$25.
This is a completely standard way for a platform to report scale to investors and the press. Amazon, Sea Limited, and every marketplace operator report GMV because it is the cleanest top-line growth story available. It also grows faster than revenue or profit in almost every case, since it is uncorrected for cost. The problem starts when the same framing gets handed to sellers as their primary success metric, through a Seller Centre home screen built around orders and sales value rather than what those orders actually returned.
The quarter the two numbers split
Sea Limited’s Q1 2026 filing is a clean, sourced example of the split, not an edge case. GAAP revenue for Shopee grew even faster than GMV, up 45.1% year on year to US$5.1 billion, which on its own would read as an unambiguous good quarter. Adjusted EBITDA, the metric closest to an actual profitability read in the filing, dropped by double digits in the same period. Sea Limited’s own 2026 guidance frames full-year adjusted EBITDA as flat to prior year in absolute dollar terms even as GMV keeps climbing toward 25% annual growth, which is the company itself stating, in writing, that the growth line and the profit line are not expected to move together this year.
None of this makes Shopee a poorly run business. Adjusted EBITDA staying flat while GMV grows a quarter of the size again is a deliberate reinvestment choice a public company can make and explain to its own shareholders. A seller cannot make the same call blind, because a seller’s version of that tradeoff is not a strategic reinvestment. It is commission, ad spend, and fulfillment fees eating into a single order’s margin without the seller necessarily noticing, because the dashboard in front of them reports the order as a win.
A SEA-specific case of what happens when nobody corrects for it
Zilingo is the sharper, more specific version of the same failure. The Singapore-based fashion marketplace raised roughly US$308 million across four funding rounds, reached a valuation near US$1 billion by 2019, and built its pitch to investors around GMV growth in exactly the way a growth-stage marketplace usually does. According to Inc42’s reporting on the company’s audited FY2019 financials, Zilingo posted losses of US$236.5 million that year alone, with cumulative losses exceeding US$430 million across 2019 to 2021. Whistleblower reports surfaced in March 2022 describing sales staff striking cash-based deals with vendors, deliberately overbilling purchases and under-invoicing sales, a mechanism that inflated the reported top line the company was raising money against. CEO Ankiti Bose was suspended within a day of the initial report and fired that May following a forensic investigation. The company entered liquidation in February 2023.
Zilingo is an extreme case, built on alleged fraud rather than an ordinary reinvestment quarter. It still illustrates the general failure mode at its clearest: a business can look large and growing by the one number everyone is watching, right up until someone checks whether the growth was ever attached to money the business kept.
Why the dashboard defaults this way
Platform Seller Centres are not built around profit because platforms do not have the seller’s cost structure. Shopee, Lazada, and TikTok Shop know the order value, the commission they charged, and the transaction fee. They do not know a seller’s cost of goods, packaging cost, or the portion of ad spend that should be attributed to a specific SKU. A dashboard that defaulted to profit would have to either ask every seller to input their own cost data or quietly guess. Guessing wrong is a worse outcome for the platform’s credibility than simply reporting the numbers it can verify: orders, GMV, and its own fee line.
That is a reasonable design constraint from the platform’s side. It is not a reason for a seller to treat GMV as the number that decides whether the quarter went well. The two questions, how much did the platform move and how much did the seller keep, require two different numbers. Only one of them appears on the home screen.
What to check instead
Contribution margin per order, revenue minus every cost that scales with that specific order, is the number GMV cannot substitute for, since it is the only one that answers whether an order made money once commission, transaction fee, payment processing, attributed ad spend, and outbound shipping are actually subtracted. A seller who wants the formula and a worked example has one available already, without needing to rebuild the math from scratch.
GMV is still worth watching. It tells a seller whether demand and reach are growing, which matters for negotiating better terms, planning inventory, and reading whether a category is heating up. It answers a different question than profit does. Treating it as an answer to the profit question is where the mistake happens, at Sea Limited’s scale or at a single Seller Centre login.
The plain answer
Track GMV to read demand and platform-level growth. Track contribution margin per order to read whether the business made money. A seller who only watches the dashboard’s default view is watching the number the platform can verify, not the number that decides whether this was a good quarter. Pull contribution margin per order at least monthly. Treat any month where GMV rises while margin per order falls as the signal to check pricing and ad spend before the next campaign, not after it.
Research Ledger (7 claims)
| Entity | Claim | Source | Status |
|---|---|---|---|
| Shopee (Sea Limited) | Q1 2026 GMV US$37.3 billion, up 30.2% year on year | Sea Limited Q1 2026 earnings release, 6-K filed with the SEC, 12 May 2026 | VERIFIED |
| Shopee (Sea Limited) | Q1 2026 GAAP revenue US$5.1 billion, up 45.1% year on year | Sea Limited Q1 2026 earnings release, 6-K filed with the SEC, 12 May 2026 | VERIFIED |
| Shopee (Sea Limited) | Q1 2026 adjusted EBITDA US$223.2 million, down 15.6% from US$264.4 million in Q1 2025 | Sea Limited Q1 2026 earnings release, 6-K filed with the SEC, 12 May 2026 | VERIFIED |
| Sea Limited | 2026 guidance: Shopee GMV to grow around 25% for the full year, full-year adjusted EBITDA no lower than 2025 in absolute dollars | Sea Limited Q1 2026 earnings release, 6-K filed with the SEC, 12 May 2026 | VERIFIED |
| Zilingo | Raised approximately US$308 million across four funding rounds, reached a valuation near US$1 billion by 2019 | Wikipedia, Zilingo entry, cross-referenced against contemporaneous funding-round press coverage | VERIFIED |
| Zilingo | Whistleblower report 30 March 2022, CEO Ankiti Bose suspended 31 March 2022, fired 20 May 2022, company entered liquidation February 2023 | Wikipedia, Zilingo entry, cross-referenced against contemporaneous Reuters and TechCrunch coverage | VERIFIED |
| Zilingo | FY2019 losses of US$236.5 million, cumulative losses exceeding US$430 million across 2019 to 2021 | Inc42 reporting, citing the company’s own audited FY2019 financials | INFERRED, Inc42’s characterization of audited figures not independently viewed |
The exact dollar figures on Zilingo’s losses come from Inc42’s reporting on the company’s audited financials, not from this site’s own review of those statements. They are flagged INFERRED on that basis. Every other claim in this piece traces to a primary filing or a cross-referenced public record.