Quarterly Statistical Report

Q1 2026 followed a familiar two-phase pattern — consistent with what was seen in Q1 2025. January to February brought a demand slowdown: order volumes and GMV both pulled back, a soft start consistent with post-holiday cooling seen in the same period last year. February to March then reversed course, with both orders and revenue recovering strongly across the panel. This dip-then-recovery shape appears to be a recurring seasonal dynamic rather than an isolated event.

Importantly, average order values stayed relatively flat throughout, down just ~3% in February and stable in March. This suggests the March surge was primarily driven by buyer expansion rather than larger basket sizes. For sellers, this is typically a more sustainable form of growth as it points to genuine demand expansion rather than a temporary price-driven spike.

Fulfillment held up well through all of this. The typical seller maintained a ~93% order completion rate across Q1, a strong baseline that suggests operational capacity kept pace even as demand accelerated in March. This is a blended figure across all marketplaces; cancellation rates vary meaningfully by marketplace, with detail in Section 4.

Three in four sellers added new listings during Q1, suggesting that catalog experimentation is a consistent behaviour across the board regardless of how demand is trending. But the data shows that how you launch matters far more than how much you launch.

New SKU success rates within the first 30 days split sharply across these two approaches. The overall median sits at 11.2%, but that single number masks the real insight. Sellers who took a curated approach, launching a smaller focused set of new products, consistently outperformed those who relied on volume alone.

A note on definition: SKU success rate here measures the share of newly listed products that generated at least one completed sale within their first 30 days of being live.

Revenue concentration across the board tells an encouraging story. The typical seller's top 10 SKUs accounted for just under half of total revenue at the start of the quarter, but that share declined to around 40% by March. As demand recovered, revenue spread more evenly across a broader set of products.

One likely contributing factor is how marketplace discovery algorithms behave during demand surges. When buyer traffic increases sharply — as it did in March — platforms surface a wider range of products across search results and recommendation feeds rather than amplifying the same top-sellers. This exposes mid-catalogue SKUs to buyers who would not have seen them in quieter months, pulling revenue away from hero SKUs organically. Sellers who had built out their catalogue depth were positioned to capture this — those still relying on a narrow set of top products had less to gain from the surge.

Year-on-year, catalog breadth is expanding: active listings grew ~25% versus Q1 2025. Yet revenue concentration is simultaneously becoming more disciplined — top 10 SKU share fell ~12 percentage points year-on-year. Notably, the new SKU success median barely moved (+0.4pp YoY), confirming that while sellers are listing more, the hit rate hasn't structurally changed. Thoughtful curation remains the differentiator, not volume.

Singapore eCommerce remains overwhelmingly a first-time buyer market. Across Q1 2026, roughly ~83% of unique buyers per seller were making their first purchase from that seller — not necessarily a new marketplace account overall, but a buyer transacting with this seller for the first time. This ratio is essentially unchanged from Q1 2025 (+0.25pp). The acquisition engine is running, but the conversion to repeat buying barely moves year after year.

Definition: A "new buyer" here means someone making their first purchase from a given seller — not necessarily someone new to the marketplace. A buyer active across multiple sellers would be counted as new for each seller they transact with for the first time.

Returning buyers punch well above their weight. While comprising just ~17% of the buyer base, they account for ~23% of total revenue — a share that grew +1.3 percentage points year-on-year. The AOV premium that returning buyers carry over new buyers narrowed meaningfully, from ~19.5% in Q1 2025 to ~14.2% in Q1 2026 (−5.3pp). This suggests new buyers' basket sizes are gradually catching up, though returning buyers still consistently spend more per order.

The conversion rate from first purchase to repeat customer sits at ~4.3% — the share of new Q1 buyers who placed a second order before the quarter ended, panel median — effectively flat versus Q1 2025 (−0.2pp). With over 80% of the buyer base being net new each quarter, this remains the single highest-leverage intervention point in seller economics. Category purchasing patterns offer no clear shortcut here: no systematic difference emerges between the categories new and returning buyers purchase from. The differentiation opportunity lies in engagement mechanics, not product assortment.

The low repeat rate is not primarily a seller execution problem — it appears to be a structural feature of how marketplace platforms are designed. Platforms are built to compete for buyer attention at the session level, not the seller level. Their discovery feeds, coin rewards, and new-user voucher programmes are designed to surface the next deal from a different seller, not to bring a buyer back to the same one. In this environment, a buyer returning to the same seller after a first purchase requires overcoming the platform's own incentive structure — which may explain why the ~4.3% repeat rate has held essentially flat for two consecutive years regardless of broader market conditions.

Customer Engagement & Chat Behaviour. Chat activity is concentrated firmly within office hours, peaking Wednesday through Friday between 9am and 6pm. This pattern points to buyers browsing and making purchasing decisions during the working day rather than during leisure time. There is a secondary spike on Friday evenings, consistent with end-of-week wind-down browsing. Weekends see significantly lower engagement across the board. Of the chat traffic that could be categorised, roughly 1 in 10 conversations relate to order and shipping queries. The remaining volume covers a broad mix of buyer questions including pricing, availability, and voucher enquiries.

Cancellation profiles differ significantly across the observed marketplace panel — Q1 2026 rates ranged from 4.4% to 16.9% — and the reasons behind them are just as telling as the rates themselves.

Year-on-year, cancellation rates improved across several marketplaces, declining by ~0.5 to 0.8 percentage points, while some held essentially flat. The improvement trend suggests sellers are adapting operationally, though the absolute spread signals that cancel risk remains highly channel-dependent.

Cancellation hotspots by product category:

Payment abandonment and change-of-mind cancellations are platform-agnostic problems — they appear across marketplaces under different labels but stem from the same underlying buyer behaviour. The categories affected are also consistent: personal care and health & beauty items tend to attract considered buyers who occasionally stall at payment, while apparel and accessories carry higher change-of-mind risk more broadly. On some marketplaces, a price-driven exit pattern also emerged, where buyers cancel after spotting a better deal elsewhere — consistent with an algorithmic discovery model that continues to surface competing products mid-consideration.

Promotional activity across Singapore's eCommerce marketplaces shifted meaningfully through Q1 2026, with attachment rates declining on every platform from January to March. The pattern is consistent rather than platform-specific, suggesting a broader normalisation of promotional intensity following the CNY campaign period.

Marketplace A.Flash sale attachment was elevated in January, with the highest concentration in health consumables, connectivity products, and personal care categories — areas where time-limited pricing aligns closely with platform browsing behaviour. By February and March this normalised to a lower baseline, consistent with the post-CNY pattern seen in prior periods. Bundle deals were the only promotional mechanic to grow consistently across all three months, with the strongest adoption in home furnishing and connectivity categories.

Marketplace B.Attachment rates and discount depth both eased through the quarter. Categories carrying the highest discount attachment across Q1 were personal care staples and workwear verticals — categories where competitive pricing is structurally expected rather than discretionary. Voucher usage pulled back broadly, though a subset of sellers maintained active programmes particularly in personal care categories.

Marketplace C.Attachment declined through the quarter. Among sellers continuing to discount, depth increased — the highest concentration of discounted volume was in apparel basics, food & snacks, and accessories. These overlap with the cancellation hotspots identified in Section 4. Discounted orders on Marketplace C cancel at approximately 7 percentage points higher rates than non-discounted orders, and order values compressed materially across the quarter. Sellers running discount programmes in these categories should weigh the volume benefit against cancel risk and margin compression rather than treating attachment rate as sufficient justification.