Trading company vs direct factory: the real cost difference for mid-market apparel brands

Every apparel brand I have worked with at the $20M to $200M range is leaving 8 to 12 points of gross margin on the table. They do not know it because the line item is hidden inside their FOB. They think they are paying a factory. They are paying a trading company that pays a factory.

In 2024 and 2025, with margin compression eating into apparel P&Ls across the category, that 8 to 12 points is not a nice-to-have. It is the difference between hitting plan and laying off your design team in Q4.

Here is what is actually happening to your money.

What a trading company actually does (and doesn't do)

A trading company is a sourcing layer. They take your tech pack, send it to one or more factories they have relationships with, get a quote, mark it up, and quote you. When the program runs, they sit between you and the factory on translation, sampling, QC, and shipment.

There are real things they do. They can be useful, especially for a brand that is too small for a factory to take seriously on its own. Specifically, a good trading company:

• Aggregates your small order with other brands' orders to hit factory MOQs
• Translates between your tech pack vocabulary and the factory's
• Manages the calendar across multiple factories if you are running a multi-category program
• Absorbs some of the QC and shipping coordination labor

What they do not do, despite the pitch:

• They do not own the factory. Their leverage when something goes wrong is the same leverage you would have, only with their margin layered on top.
• They do not control quality at the source. They inspect at the end.
• They do not actually move faster than direct. They move slower, because every email between you and the factory passes through them.
• They do not protect you on tariffs, compliance, or audits. You still own all of that.

The pitch is "we handle the complexity of China for you." The reality, for any brand above $20M, is that you are paying a tax to avoid building a capability you should have built two years ago.

Where the 15-25% markup comes from, line by line

Industry-standard trading company markups sit in the 15 to 25% range on FOB. Sometimes higher on small orders, sometimes lower on huge ones. Here is how that markup breaks down on a typical knit program at a $50M brand.

Assume an FOB factory price of $6.00 per unit on a basic knit top, 500,000 units a year across styles.

• Factory true cost (what the factory would quote you direct): $6.00 per unit
• Trading company "service fee" or commission: $0.60 to $0.90 per unit (10 to 15%)
• Quality and inspection layer (often padded): $0.15 to $0.30 per unit
• Shipping and consolidation markup: $0.10 to $0.20 per unit
• Sample and development fees: spread across the program, $0.05 to $0.15 per unit
• Currency and payment terms padding: $0.05 to $0.10 per unit

You land at roughly $7.20 to $7.65 per unit. On 500,000 units a year, that is $600,000 to $825,000 of margin that did not have to leave your business.

Scale it up. At $50M topline with apparel-typical COGS structures, the trading layer is eating $7M to $12M of gross margin annually across your full assortment. That number is not exaggerated. It is the average for mid-market brands I have audited.

The trading company markup is not a fee. It is a tax on not building your own sourcing capability.

The hidden costs that don't show on the quote

The line items above are the visible markup. The hidden costs are worse, and they almost never show up in a side-by-side quote comparison.

Sample round time. Trading companies run 4 to 8 weeks per sample round because your tech pack is being translated, sent, translated back, and reviewed by an intermediary. Direct factories run 2 to 4 weeks. On a six-style seasonal program with three sample rounds, that is six to twelve extra weeks. Six to twelve weeks is the difference between a Q2 launch and a Q3 launch. It is also the difference between full-price sell-through and a markdown.

Lead time padding. Trading companies quote longer factory lead times than the factory would quote direct. They need the slack because they are juggling multiple clients' programs against the factory calendar. Your "12 week" lead time was probably an 8-week factory lead time plus four weeks of trading-company slack.

Quality drift. When something goes wrong at the factory, the trading company has a financial incentive to ship it anyway and hope you do not notice. The factory does not have that incentive. The factory has a direct relationship to protect.

Tariff exposure. Trading companies are not on the hook for HTS classification mistakes. You are. If they classified your goods incorrectly to soften the quote and CBP catches it, you pay the back duty plus penalties.

No leverage on rework. When you find a defect in a finished shipment, the trading company is your only point of contact, and they are now the one negotiating with the factory on your behalf. They will pick the cheaper resolution for them, not the right one for you.

No insight into raw material. You cannot ask a trading company about yarn supplier diversification, dye lot consistency across colorways, or fiber traceability for compliance reasons. They do not know, and the answer they give you is whatever the factory told them.

Add it all up and the real cost of a trading company is closer to 25 to 35% above direct factory pricing once you include the lost margin from delayed launches and the quality risk.

Why going direct isn't simple (and how to do it without the risk)

If direct is so much cheaper, why does anyone use a trading company? Because going direct is hard. Specifically:

• You need someone bilingual on your side of the relationship who can manage the day-to-day with the factory in Mandarin or who has a counterpart at the factory who can handle English at an operational, not just sales, level.
• You need to be large enough that the factory takes you seriously. Below 3,000 units per style per colorway, most large factories will not prioritize your account.
• You need to handle your own QC. Either you fly in, you hire a third-party inspection firm, or the factory has an accredited in-house lab and you trust it.
• You need to handle your own shipping consolidation, customs paperwork, and tariff classification. None of this is hard, but it is real work.

The honest answer is that "direct" does not have to mean "do it all yourself." It means cutting out the trading company by working with a factory that has built the infrastructure to serve Western brands directly. A factory that has:

• A US-based or US-fluent counterpart on the brand-facing side
• An accredited in-house lab so QC happens at the source instead of after the fact
• The willingness to quote in 72 hours instead of three weeks
• Real vertical integration across yarn, fabric, cut and sew, and finishing so there is one accountable party for the whole program

That model is what we built at Ohzehn. Not because we invented it, but because it is what direct sourcing actually looks like when the factory does the work to be accessible to a Western brand. You get the cost structure of direct without the operational tax of running China yourself.

When a trading company still makes sense

I am not going to tell you trading companies are always wrong. They are wrong for most brands above $20M, but they have a real place. Use a trading company when:

• You are under $5M in revenue and you genuinely cannot hit factory MOQs on any style. Aggregation has real value here.
• You are running a tiny number of styles in extremely small quantities for a test or a capsule, and you do not want to build sourcing infrastructure for it.
• You are sourcing across a wide range of categories where no single factory can serve all of them and you do not have the bandwidth to manage four factory relationships.
• The trading company in question is genuinely additive: a real product developer who is shaping your tech packs, pushing back on construction, and adding value beyond the markup. Those exist. They are rare. They are worth their fee.

For everyone else, especially brands above $20M where every margin point compounds across a meaningful unit volume, the math is brutal and the answer is direct.

The decision framework

When you sit down to evaluate this, the question is not "trading company or direct." The question is: at our current scale, what is the all-in cost of the trading layer, including hidden costs, versus the all-in cost of building a direct factory relationship?

For a brand doing $20M, the trading tax is roughly $2M to $4M a year. The cost of switching, if you pick the right direct partner, is maybe $50K to $150K in transition work and a season of dual-tracking. The payback period is less than four months.

The brands that win the next five years in apparel are the ones who cut out the layers their P&L cannot afford. The brands that lose are the ones still paying a 20% tax to avoid a conversation in Mandarin.

The conversation is not actually that hard. The 20% is.

Dougie Taylor

Co-Founder, Ohzehn Textiles · Forbes & Inc. recognized brand operator