Last year, a buyer we work with placed a $47,000 order for LED panel lights. The supplier confirmed. The contract was signed. Four weeks later — three days before the ship date — the supplier called. "We need two more weeks."
Two weeks became four. Four became six. The buyer's client canceled. $47,000 in product sat in a Shenzhen warehouse for three months before finding a new buyer at a 30% discount.
The factory hadn't lied about their theoretical capacity. They had the machines. They had the line count. What they didn't have was available capacity — three other large orders were running simultaneously, and the buyer never asked.
Here's the thing: most B2B capacity failures aren't about factories being too small. They're about factories being too busy — and buyers not having a framework to detect it.
We've built this framework from data on 2,300+ supplier profiles on Compare2Best, cross-referenced with actual delivery performance. Five dimensions. Each one catches something different.
Start here. Ask for the number of assembly lines, SMT lines (for electronics), and packaging lines. Then ask the uncomfortable follow-up:
"How many lines are currently running? At what shift pattern?"
If they claim 8 production lines but only 5 are running — and it's not a weekend or holiday — something's off. Either demand is weak (which means the business might be unstable) or they're hiding downtime. Neither is good.
Do the math: lines × output per line per shift × shifts per day × working days per month. Then multiply by 0.7. That's the realistic number. If your monthly order is 60,000 units and their realistic capacity is 70,000, the buffer is just 10,000 units. One machine breakdown and you're late.
A factory with 100% capacity but two weeks of raw material inventory is a factory that will stop. Cold.
Walk the warehouse. Look at the material shelves. Are they full or sparse? Check the dates on incoming material logs. A supplier running at genuine volume has a steady stream of material deliveries — you'll see truck schedules, receiving logs, and supplier invoices that tell the story.
Ask: "What's your buffer stock policy for critical components?" For LED lighting, that means LED chips, drivers, aluminum housings. If they keep less than 3 weeks of buffer stock on their top 3 components, one supply chain hiccup cascades into your delayed order.
Machines don't assemble products. People do. And in manufacturing hubs like Zhongshan and Ningbo, workforce churn can hit 30–40% annually.
Red flags we've catalogued:
Ask for the worker attendance record for the past 3 months. Not names — just headcount per day. A factory producing consistently shows flat or gently rising headcount. A factory with capacity problems shows spikes — temp workers hired for specific orders, then let go.
Every factory has a "best month." Ask for it.
"What was your highest-volume month in the past year? How many units did you ship?"
If their peak month was 80,000 units and your order is 70,000/month, you're asking them to run at 87.5% of their all-time best — for multiple consecutive months. That's a red flag unless they've added capacity since then.
Also ask: "Can I see shipping records from that month?" A factory that genuinely shipped 80,000 units has bills of lading, container bookings, and freight forwarder confirmations. If they hesitate, the number is inflated.
Old machines break. New machines have teething problems. The sweet spot for most light-industrial equipment is 2–7 years old — past the infant mortality phase but before wear-related failures spike.
On a site visit, look at the maintenance log. Not the shiny one they keep for audits — the greasy one on the workshop floor. Check:
A factory running machines past year 10 without a documented maintenance schedule is a factory that will have a breakdown during your production run. It's not a question of if. It's a question of which week.
| Dimension | 🔴 Red Flag | 🟢 Green Flag |
|---|---|---|
| Line Utilization | Claimed 10 lines, only 6 running on a weekday | All lines active; can show last 30 days of daily output logs |
| Raw Materials | Warehouse shelves sparse; <2 weeks buffer stock | 4+ weeks of critical components; visible supplier delivery schedule |
| Workforce | Headcount graph shows sharp spikes and dips | Flat or gently rising headcount; workers can explain QC steps |
| Peak History | Can't or won't share peak month shipping records | Provides bills of lading from their best month within 24 hours |
| Equipment | Machines 10+ years old, no maintenance log available | 2–7 year equipment age; greasy maintenance log on shop floor |
Compare verified supplier profiles with documented production capacity data
This guide is produced by the Compare2Best knowledge team and reviewed by B2B procurement specialists. Data cited from supplier profiles and delivery performance records on the Compare2Best platform.