全球采购中的MOQ正在改变:小批量、多SKU需求如何重塑制造模式?

MOQ不会消失,真正发生变化的是:
**MOQ正在从一个固定数字,变成与材料、SKU、换线、包装和订单规模联动的商业模型。**
中国智能制造基础仍在继续扩大。官方数据显示,截至2025年初,中国已建设超过3万家基础级智能工厂、1,200家先进级和230家卓越级智能工厂。
数字化和自动化提高了制造柔性,但并不意味着“小订单没有额外成本”。
一、为什么MOQ存在?
生产一个产品通常需要:
- 调机;
- 换模;
- 采购材料;
- 首件确认;
- 检验;
- 包装。
这些准备成本不会随着订单数量线性下降。
因此:
数量越小,每件产品承担的准备成本越高。
二、多SKU正在改变传统MOQ
例如:
- 5种颜色;
- 4种尺寸;
- 3种包装;
就形成60个SKU。
如果每个SKU MOQ都是1000件:
总订单=60,000件。
客户可能根本不需要这么多库存。
因此采购正在增加两个概念:
**Total Order MOQ**
和
**Per-SKU MOQ**
例如:
总MOQ 10,000件
每SKU最低300件
这比单纯规定每个SKU 5,000件更适合多品种业务。
三、小批量最大的成本是Changeover
柔性制造的难点通常不是:
“机器能不能生产?”
而是:
- 换模需要多久;
- 换色需要多久;
- 材料MOQ是多少;
- 包装MOQ是多少;
- 每次首检成本是多少。
因此小批量生产首先是生产组织能力。
四、Stellar四层MOQ模型
Sample MOQ
用于样品和工程验证。
Market Test MOQ
用于市场测试。
Standard MOQ
进入常规生产经济区间。
Volume Tier
规模生产价格区间。
例如:
| 数量 | 场景 |
|---|---|
| 100 | Sample |
| 500 | Market Test |
| 2,000 | Standard |
| 10,000+ | Volume |
不同层级对应不同单价。
五、买家应该如何谈MOQ?
不要简单问:
Can you lower your MOQ?
更有效的是告诉供应商:
- 首批数量;
- 年度预测;
- SKU数量;
- 后续订单计划。
然后谈:
**阶梯MOQ+阶梯价格。**
---
MOQ真正管理的是:
**制造成本、买家库存风险和市场验证风险之间的平衡。**
未来更有竞争力的工厂,不一定MOQ最低,而是能够支持:
**Sample → Small Batch → Multi-SKU → Mass Production**
平滑扩张。
MOQs are not disappearing; what is truly changing is this:
**MOQs are shifting from a fixed number to a business model linked to materials, SKUs, changeovers, packaging, and order volume.**
China's foundation for smart manufacturing continues to expand. Official data shows that as of early 2025, China has established over 30,000 "basic-level" smart factories, 1,200 "advanced-level" smart factories, and 230 "excellence-level" smart factories.
While digitalization and automation have increased manufacturing flexibility, this does not mean that "small orders come without extra costs."
I. Why do MOQs exist?
Producing a product typically requires:
- Machine setup;
- Mold changes;
- Material procurement;
- First-article inspection/approval;
- Quality inspection;
- Packaging.
These setup costs do not decrease linearly with order volume.
Therefore:
The smaller the quantity, the higher the setup cost borne by each unit.
II. Multi-SKU scenarios are changing traditional MOQs
For example:
- 5 colors;
- 4 sizes;
- 3 packaging types;
...results in 60 SKUs.
If the MOQ for each SKU is 1,000 units:
Total order = 60,000 units.
The customer may not need that much inventory at all.
Consequently, procurement strategies are increasingly incorporating two concepts:
**Total Order MOQ**
and
**Per-SKU MOQ**
For example:
Total MOQ: 10,000 units
Minimum per SKU: 300 units
This approach is better suited for high-variety businesses than simply mandating an MOQ of 5,000 units per SKU.
III. The biggest cost in small-batch production is the changeover
The challenge in flexible manufacturing is usually not:
"Can the machine produce it?"
But rather:
- How long does a mold change take?
- How long does a color change take?
- What is the material MOQ?
- What is the packaging MOQ?
- What is the cost of each first-article inspection?
Therefore, small-batch production is primarily a test of production organization capabilities.
IV. Stellar's Four-Tier MOQ Model
Sample MOQ
For samples and engineering validation.
Market Test MOQ
For market testing.
Standard MOQ
Entering the economic range for routine production. ### Volume Tier
Price ranges based on production scale.
For example:
| Quantity | Scenario |
|---|---|
| 100 | Sample |
| 500 | Market Test |
| 2,000 | Standard |
| 10,000+ | Volume |
Different tiers correspond to different unit prices.
V. How Should Buyers Negotiate MOQ?
Don't simply ask:
Can you lower your MOQ?
It is more effective to provide the supplier with:
- Initial order quantity;
- Annual forecast;
- Number of SKUs;
- Follow-up order plan.
Then negotiate based on:
**Tiered MOQ + Tiered Pricing.**
---
What MOQ really manages is:
**The balance between manufacturing costs, buyer inventory risk, and market validation risk.**
In the future, the most competitive factories won't necessarily be those with the lowest MOQs, but rather those capable of supporting:
**Sample → Small Batch → Multi-SKU → Mass Production**
...and enabling smooth, scalable growth.