When brands calculate their sourcing cost, they typically look at the FOB price on the invoice. What they don't see is the 5–12% of landed cost that disappears into the agent model before the invoice is even issued.
This is not an accusation of dishonesty. It is a structural feature of how traditional sourcing agents are incentivised - and understanding it is the first step to eliminating it.
The agent model exists for a real reason: before digital tools made factory information, compliance status, and pricing transparent, brands genuinely needed a local intermediary with relationships and language capability to place orders at all. That value was real. But the commission-plus-rebate structure that funds the agent's business was built for an era before brands had any way to verify what was happening underneath the invoice - and the incentive misalignment it created has persisted long after the information gap that justified it closed.
In our experience auditing traditional sourcing programmes, hidden costs in a traditional agent model typically run 22-24% of the visible FOB value - undisclosed factory rebates (2-5%), buffer overordering (5-10% extra units), opaque fabric margins (8-15%), and excess sampling rounds. On a $40,000 order, that's roughly $8,700-9,600 that never appears itemised on any document the brand sees. At $5M FOB per year, it's $300,000-600,000 annually. None of it requires dishonesty to occur - it's a structural feature of how commission-based agents are paid, not a one-off.
Where the hidden cost lives
1. The agent commission and rebate structure
Most sourcing agents charge a visible commission - typically 5–8% of FOB value. What is less visible is the factory rebate: factories pay agents a separate percentage (often 2–5%) for bringing them business. This rebate is not disclosed to the brand, and it creates an incentive for the agent to direct orders to the factories that pay the highest rebates - not necessarily the factories that produce the best quality at the best cost.
2. Buffer overordering
Agents frequently recommend ordering 5–10% more units than needed "to cover quality rejections." On a 5,000-piece order at $8 FOB, that is an additional $2,000–4,000 of inventory that will either be marked down, stored, or discarded. The agent's commission applies to the full inflated quantity.
3. Opaque fabric sourcing margins
In many agent-managed programmes, the agent also sources fabric. The mill price is known to the agent; what the brand pays for the fabric is a separate negotiation. Margins of 8–15% on fabric are common and almost never disclosed.
4. Sampling over-iteration
Poorly managed sampling processes - common in agent-run programmes with no version control - result in 4–6 rounds of sampling where 2–3 would suffice. Each round costs $200–600 and 10–14 days. Across a full season's development, this adds $5,000–15,000 in direct costs and 6–8 weeks in time-to-market.
A worked example on a real order size
Take a 5,000-piece style at $8 FOB through a traditional agent - a $40,000 order on paper. Layer in what typically happens underneath:
- Visible agent commission at 6%: $2,400
- Undisclosed factory rebate at 3%, paid separately by the factory and baked into the FOB the brand never sees the breakdown of: roughly $1,200
- Buffer overordering at 7% (350 extra pieces the brand did not need): $2,800
- Fabric margin at 10% on a programme where the agent also controls fabric sourcing: $1,500–2,000
- Excess sampling rounds (5 rounds instead of 2–3, at $400 average per round): $800–1,200
Total hidden cost on this single style: roughly $8,700–9,600 against a $40,000 nominal order - between 22% and 24% of the visible FOB value, most of it never itemised on any document the brand actually sees.
Scale that across a full season of 15–20 styles and the number stops being an abstraction. It becomes the difference between a sourcing programme that supports healthy gross margin and one that quietly erodes it every quarter - money that shows up nowhere on an invoice, but shows up every time on the P&L.
How to audit your current agent relationship
You do not need to switch sourcing partners to find out whether this applies to you. A few direct questions will surface most of it:
- Ask for the fabric mill invoice directly, not the agent's fabric cost line. If the agent controls fabric sourcing and cannot produce the underlying mill invoice, there is likely an undisclosed margin.
- Ask whether the factory pays the agent anything beyond what you pay the agent. Most agents will not volunteer this, but a direct question about factory-side compensation often gets a more honest answer than expected.
- Compare your ordered quantity against your actual received quantity over the last four seasons. A consistent pattern of receiving 5–10% more than the PO quantity, with a "quality buffer" explanation, is the overordering pattern described above.
- Count your sampling rounds per style. More than three rounds to reach an approved sample, consistently, points to a process problem that is costing you money and time regardless of intent.
None of this requires an adversarial conversation. A sourcing partner with nothing to hide will answer these questions directly. One that deflects or cannot produce documentation is telling you something important.
The transparent alternative
Operator-led sourcing with open-book costing looks different. Every cost line is visible: fabric rate per kilogram, CMT per piece, trims, freight, Tradio's fee. The fee is fixed and disclosed upfront. There are no factory rebates because we do not earn from factory placement - we earn from programme outcomes.
The quality guarantee structure reinforces this. When we absorb the cost of quality rejections (our policy on failures that pass our QC process), we have a direct financial incentive to get QC right the first time. That incentive alignment is absent in a commission-based agent model where the agent is paid regardless of rejection rate.
The math is not complicated. A mid-size brand sourcing $5M FOB per year through a traditional agent is likely paying $300,000–600,000 in hidden costs annually. Eliminating those costs - or even halving them - is a significant business outcome that shows up in gross margin, not just sourcing spend.
If you want to map what that looks like for your specific programme, request a sourcing audit. And if part of your exposure is EU-bound, worth reading alongside this: our EUDR compliance checklist - agent-managed programmes are the most common source of the compliance data gaps we see.
FAQ
Is a sourcing agent commission always a red flag?
No - a disclosed, fixed commission is a legitimate and common model. The problem is undisclosed layers stacked on top of it: rebates, fabric margins, and overordering that the brand never sees itemised.
How do I know if I am being overordered on purpose?
Track received quantity against ordered quantity over several seasons. A one-off variance can be a genuine yield issue; a consistent 5-10% pattern across styles and seasons is a structural buffer, not a coincidence.
Will switching to open-book costing slow down my sourcing process?
Generally the opposite - transparent costing removes the negotiation friction that opaque margins create, since there is nothing to negotiate around. Most brands find the process faster once every cost line is visible upfront.
Does open-book costing mean I lose the agent's factory relationships and leverage?
No. An operator-led model still maintains direct factory relationships and negotiating leverage - the difference is that the incentive behind those relationships is aligned to your outcomes, not to a hidden commission structure.
What is a reasonable sourcing fee to expect under a transparent model?
Fixed fees in an open-book model typically land in a similar range to a traditional agent's visible commission, but without the additional 15-18 percentage points of hidden cost layered on top - the total cost to the brand is materially lower even though the disclosed fee looks comparable on paper.