The short version. A trading company sells you a price. A factory sells you a process. Both can make a good cap. Only one of them can tell you why the second half of your order came out a different shade of navy.
Why buyers ask this. Most custom headwear failures are not design failures. They are interpretation failures: the quote said "navy", the sample looked right, the bulk came out 4% darker, and nobody can prove who changed what. That gap opens wider with every layer between you and the sewing floor.
What a trading company actually does. Collect an enquiry, forward it to one of several factories, mark up the reply, and coordinate. This is not dishonest work. It has real value when you need twenty product categories from one invoice. But it introduces a translation layer: your specification becomes their specification becomes the factory's specification.
Where the money goes. The markup is visible in the unit price, so buyers usually spot it. What is less visible is the cost of ambiguity: resampling after a misunderstood detail, colour matching done twice, an inspection that passes a standard nobody wrote down. Those costs land on you, not on the trading company.
Five ways to tell them apart in ten minutes. 1. **Ask which legal entity manufactures the goods.** A factory answers with its own name and site. A trader answers with a partner's name, or hedges. 2. **Ask what happens if the bulk fails inspection.** A factory can rework on its own floor. A trader has to negotiate with a third party whose margin is already fixed. 3. **Ask to see the line making something comparable.** Not a showroom, not a sample wall — production. 4. **Ask who owns the audit.** Social audits (BSCI, SMETA, Better Work) are issued against a legal entity and a site. If the certificate belongs to a company you have never heard of, you now know the structure. 5. **Ask for the production plan in writing.** A factory can commit to one. A trader can only relay one.
The honest counter-argument. Own-factory suppliers are not automatically better. A small factory with one line and no compliance function will struggle with a licensed retail programme. A well-run trading company with audited partners can outperform a poorly managed factory. The question is not which model is superior — it is whether the capability matches your programme.
Where an own-factory model wins clearly. - **Licensed and retail programmes.** Retailers audit the site, not the invoice. Owned sites mean the audit is on your manufacturer. - **Reproduction across repeat orders.** Same floor, same equipment, same people — variance is a process problem you can fix. - **Speed on simple builds.** No relay means a 50-piece run with stock materials can move immediately rather than waiting for a third party to schedule it. - **Flexible capacity across sites.** Multiple owned sites let volume move when one is loaded.
Where a trading company can win. Breadth of categories, single-invoice convenience, and lower minimums on niche items a factory does not specialise in. If your programme is wide and shallow, that can be the right call.
A practical test to run this week. Send the same written specification to three suppliers — one obvious trader, one obvious factory, one you are unsure about. Ask each the five questions above. Compare not the prices but the specificity of the answers. The supplier who writes back with questions about your crown structure and closure type is the one who will make your caps.
What Gazelle is, plainly. Three wholly owned factories in Dongguan, Suzhou (Anhui) and Phnom Penh. More than ten million pieces a year. 800+ people. The factory that quotes your order is the factory that makes it. If that is not what your programme needs, we will tell you rather than quote anyway.