**To:** Priya Raman
**From:** Lena Hartmann
**Re:** Fulfillment operational health — first-week findings

**1. End-to-end delivery time (priority #1).** Typical delivery (order placed → delivered) runs 3.6–4.3 days on average, median 3–4 days, with 83–90% of orders inside 5 days in a normal month (monthly trend report, 2024–2025). That's held roughly flat over the past two years — no structural drift, good or bad. The real variability is between merchants, not over time: 2025 pct-within-5-days ranges from 92% (Pfotenglück, Cragsley Outfitters) down to 54–58% (Thistlecombe, Purrfect Pantry) — genuinely a coin toss for some. Critically, this doesn't track vertical (Food & Coffee has both the best and some of the worst performers), so it isn't a category problem.

**2. Peak season (priority #2).** Nov–Dec 2025 is the stress test: weekly data shows pct-within-5-days falling to 73.5% (week of Nov 24) and 56.1% (week of Dec 1) as volume roughly doubled, with backlog (undelivered orders) tripling. It recovered to 80–93% by mid-to-late December. Decomposing the time (processing vs. transit, monthly) shows warehouses held up fine — processing time stayed flat at ~1.1–1.2 days straight through peak — so the slowdown was entirely in carrier transit (2.5–2.9 days baseline → 3.46 in November). **The warehouses are not the bottleneck; carriers are.**

**3. Weak links (priority #3) — high confidence.** Two carriers are chronically weak all year, not just at peak: Landfall Freight (13.6% within 3 days in Q4, similarly bad all year) and Lowlands Express (44.6%). A third, Albion Parcel, was solid through October (75.7% within 3 days) then collapsed in November (28.2%, avg 7.5 days) and got *worse* in December (28.6%, avg 9.7 days, p90 ~21 days) — no recovery by year-end. Together these three carried ~32% of Q4 volume. Most importantly, carrier-mix data directly explains the merchant variance from #1: the same merchant shows wildly different reliability by carrier (e.g., Grind Collective: 88.7% within 3 days on Kennet Freight vs. 19.4% on Landfall Freight). The "bad" merchants are simply the ones with meaningful volume stuck on weak carriers — this is a carrier issue, not a merchant or warehouse issue.

**4. Replenishment (priority #4) — partial, lower confidence.** Across all suppliers, 40–85% of purchase orders arrive after their expected date, and this rate (47–58%) is stable across quarters — it's a chronic baseline issue, not peak-specific. A few supplier families ("Roasters," "Textiles" from certain vendors) show the longest promised lead times (28–43 days) and the worst lateness. The Q4 "open PO" count (418, vs. 0 in Q1–Q3) looks alarming but I don't trust it as a standalone signal: POs placed late in December haven't had time to be received before our 2025-12-31 cutoff, so this is partly a right-censoring artifact rather than proof of a real backlog — needs a cleaner test (e.g., restrict to POs ordered before Nov 15). Current stockout exposure (point-in-time snapshot only, not a trend) ranges from <2% to ~15% of tracked SKUs out of stock by warehouse, with Köln (Brennquell Kaffee) highest; I'd treat this as a rough baseline, not a reliable historical pattern, since we have no stockout event history to check if it's worse than usual.

**Caveats:** delivery/transit definitions require all shipments on an order to be fully delivered, so partially-shipped orders are excluded; very recent weeks/months understate reliability simply because deliveries are still in transit at the data cutoff (flagged in each affected report).

**Next steps:** (1) a clean, non-censored read on the Q4 PO backlog; (2) confirm whether Albion Parcel and Landfall Freight are concentrated on specific lanes/regions before the carrier conversation; (3) returns/restock cycle time, deferred this week as agreed.