5 E-commerce Operations Lessons from a Near-Collapse on Singles' Day

Last Singles' Day, at 3 a.m., I crouched by the warehouse door, staring at piles of unsent packages and inventory that didn't match the system. I wanted to kick myself. At 10 a.m., my customer service lead messaged: "Boss, another 20 customers requested refunds, claiming we faked shipment." I was numb. That night, I lost three long-time clients and paid 8,000 yuan in penalties.

TL;DR That boom nearly broke me, but it also woke me up. Over the next six months, I overhauled everything from inventory management to shipping processes. Here are the hard-earned lessons from my e-commerce operations journey.

Inaccurate Inventory? That's a Time Bomb

Before Singles' Day, I told my ops team confidently, "Inventory is fine—5,000 units in the system." Two hours after sales started, the system showed 2,000 oversold. Customers bombarded us with complaints, and I had to call each one to apologize. Later I found the root cause: during receiving, workers skipped scanning individual units, so the system numbers were already wrong.^[1]

The root cause of inventory inaccuracy isn't a bad system—it's a flawed process. From then on, I mandated scanning every single unit during receiving.

Standardize Receiving Process

Previously, receiving was done by "eyeballing"—workers glanced at box labels and entered data. Now I set a hard rule:

After the change, inventory accuracy jumped from 75% to 99%.^[2] Numbers don't lie.

Cycle Counts Beat Monthly Full Counts

I used to rely on monthly full counts, which took a whole day and left employees grumpy—yet data still didn't match. I switched to weekly cycle counts covering 20% of SKUs. In a month, every item gets checked without disrupting daily work.

Slow Shipping? Unclog the Information Flow

On Singles' Day, our shipping process was: ops export orders → print labels → manual sorting → packing → wait for courier pickup. Every step was manual, painfully slow. The worst: six hours from order to dispatch.^[3]

The bottleneck isn't manpower—it's information flow. I integrated ERP with WMS; orders now auto-push to warehouse PDAs, and pickers scan to confirm.

Optimize Picking Routes

Pickers used to run all over the warehouse, clocking 20,000 steps a day with pitiful efficiency. I rearranged bins, placing hot sellers near the packing zone.

Results speak: we shipped 95% of orders on Singles' Day after optimization.

Batch Picking Beats Single-Order Picking

Previously, pickers grabbed one order at a time, running back and forth. Now I group orders by zone and category, picking a batch before packing. Efficiency doubled.

Returns? Don't Let Them Become a Mess

The post-Singles' Day returns wave was a nightmare. Returned items piled up in a corner, ignored. By the time we processed them, some had passed the return window, others were swapped.

Returns aren't a cost center—handled well, they become a profit center. I set up a dedicated returns team with a "three-hour rule": all returns must be inspected and restocked within three hours of arrival.

Tiered Returns Processing

This cut return turnaround from two weeks to two days, and the clearance section's revenue offset some return costs.

Data-Driven? Don't Just Watch GMV

I used to check only three metrics: sales, order count, refund rate. On Singles' Day, sales hit a record, but I lost money. Now I track finer metrics: average order value, unit price, inventory turnover days, fulfillment cost ratio.

The truth of operations hides in detailed data. For instance, one hot product had a 30% return rate. Investigation revealed a sizing label error. After correction, returns dropped to 8%.

Watch Fulfillment Cost Ratio

I never calculated this before, thinking sales were all that mattered. Later I found fulfillment (storage + packing + shipping) ate 18% of revenue, well above the industry average of 12%. After optimization, it dropped to 11%, saving nearly 200,000 yuan a year.

Team Management? Don't Treat People Like Machines

During Singles' Day, employees worked overtime for days, exhausted. A young woman cried quietly at night; I saw her. Next morning, I bought breakfast, held a quick meeting, and said, "It's okay if orders aren't shipped—people come first."^[4]

The bottleneck in team efficiency is often emotion, not skill. I adjusted shifts, set up "energy stations," and brought in part-timers during peak seasons.

Instant Rewards

Bonuses used to come at year-end—employees had forgotten what they did. Now I give weekly awards: 500 yuan cash to the most efficient team, handed out on the spot. It worked wonders for morale.

Conclusion

Looking back at last year's Singles' Day disaster, my biggest takeaway is: e-commerce success isn't about one hot product—it's a system of inventory, processes, data, and team. After stumbling, I built Flash WMS to embed these lessons into a system.

Key Takeaways:

• Inaccurate inventory? Fix with scanning and cycle counts
• Slow shipping? Unclog information flow, optimize picking
• Returns piling up? Set up a dedicated team, process within 3 hours
• Only watching GMV? Track fulfillment cost and return rate
• Team burnt out? Give instant rewards, put people first

---

References

1. Fortune Business Insights WMS Market Report — Cited for importance of inventory management
2. Gartner Supply Chain Research — Cited for inventory accuracy improvement data
3. McKinsey Operations Insights — Cited for shipping process optimization data
4. China Federation of Logistics & Purchasing — Cited for team management best practices