The overseas middleman trap, real cases, and how to reach buyers directly

For decades, the path out of China looked the same. A factory in Zhejiang or Guangdong made an excellent product. An overseas trading agent put their name on the relationship. The buyer in Hamburg or Houston never learned where the goods actually came from. The factory got an order. The agent got the customer. Guess which one is worth more.

That arrangement made sense when a manufacturer in Yuhuan had no realistic way to find, vet, or talk to a distributor 12,000 kilometers away. It makes far less sense now. China's cross-border e-commerce exports grew nearly 17% year-on-year in 2024, reaching roughly 2.15 trillion RMB, according to China's customs data reported via the State Council. The infrastructure to reach buyers directly now exists. The habit of routing everything through a middleman has not caught up, and that gap is exactly where margin quietly disappears.

This guide is long on purpose. The middleman trap is not one mistake you can fix in an afternoon. It is a structure, built over years, that quietly hands your most valuable asset to someone else. To dismantle it you need to understand how the structure works, why it persists, what the real cost is in numbers, who has broken out and how, and the precise sequence of moves that puts you back in control. We will walk through all of it: the economics, the real cases, a step-by-step playbook, the pitfalls that sink most attempts, the metrics that tell you whether you are winning, and a long FAQ. Read it once end to end, then keep it as a reference.

One framing to hold onto before we begin. The single most valuable thing in any export business is not the factory, the tooling, or even the product. It is the relationship with the person who decides to buy. Machines depreciate and products get copied, but a relationship with a buyer who trusts you, knows your name, and calls you first compounds in value every year. The middleman trap is, at root, a slow transfer of that one asset out of your hands and into someone else's. Everything that follows in this guide is about getting it back, or building it from scratch if you never had it. Keep that lens on as you read: at each step, ask who ends up owning the buyer relationship. If the answer is anyone but you, the trap is still closed.

What the middleman trap actually is

Most exporters picture a "middleman" as one person: a broker, a trading company, an agent who introduced them to a customer years ago and has skimmed a commission ever since. That picture is too small. The trap is not a person. It is a position in the value chain, and whoever holds that position holds the leverage.

The chain, and where you sit on it

A typical export chain has six links: the Chinese factory, a Chinese export company, an overseas importer, an overseas wholesaler, an overseas retailer, and finally the end customer. The factory carries the heaviest risk in the whole chain. It bought the tooling, hired the workers, financed the raw materials, and absorbed every swing in steel, resin, or copper prices. Yet it sits at the very bottom of the value ladder, capturing the thinnest slice of the final price. Every link above it adds a markup the end buyer pays but the factory never sees.

The longer the chain, the further you are from the price the market is actually willing to pay, and the weaker your negotiating position becomes. This is not bad luck or a tough customer. It is the structural destiny of traditional export trade. If you have ever felt that you work the hardest and earn the least in your own supply chain, the chain itself is the reason, not your sales skills.

Three gates the middleman controls

The intermediary does not simply take a cut. It sits on top of three gates, and as long as it controls any one of them, you are exposed.

The margin gate. Each hop compresses your price. The broker represents their own commission first, and will happily undercut your price to close a quick deal, because their incentive is volume of transactions, not the long-term value of your brand.

The information gate. No direct contact means no feedback. You do not know which SKUs are growing, what the end buyer's customers complain about, what a competitor is charging, or why an order suddenly dropped by 40%. You are manufacturing in the dark, optimizing for a purchase order instead of for real demand.

The relationship gate. The contract, the payment terms, and the loyalty all sit with the intermediary. You are interchangeable. The day a cheaper factory appears, the agent switches suppliers and you lose an account you never really had, because it was never really yours.

The platform version of the problem (where a marketplace becomes the middleman) follows the same logic; we cover it in brand versus marketplaces and in the cost breakdown of TikTok Shop going global. The mechanics rhyme: someone else owns the customer, and that someone else sets your ceiling.

The real cost, in numbers

Founders nod along to "the middleman takes margin" and then keep operating exactly as before, because the cost feels abstract. It is not abstract. Let us put illustrative numbers on it so you can run the math against your own product. The figures below are typical, framed as illustration, not precise claims about any single company.

Where the price goes on a single unit

Take a product that leaves your factory at a 100 RMB equivalent ex-works price. Here is how a long intermediated chain and a direct model can split the final price the end buyer pays. Treat these as illustrative ranges, not fixed laws.

The lesson is not "everyone above you is greedy." Distributors and retailers earn their margin by carrying inventory, extending credit, and owning shelf space. The lesson is that in the intermediated chain, the factory captures a minority of the value it created, while in a direct model the same factory keeps the majority. That spread is your reinvestment budget, your R&D budget, and your survival cushion when prices get cut.

Run the math on your own product before you read further, because the abstract version of this argument never changes behavior. Take your real ex-works price. Find out, even roughly, what the same item retails for in your largest export market. Divide the first number by the second. If you are capturing less than half of the final price, you are funding a chain of intermediaries with margin you could be keeping or reinvesting. That single ratio is often the most clarifying number a founder runs all year, because it converts a vague grievance into a budget line. Once you see that a 10-point shift in your captured share could double your R&D spend, the priority sorts itself out.

There is a second-order effect worth naming. When you capture more of the price, you also gain the freedom to not compete on price. A factory living on a 25% share has no room to invest in quality, certification, or service, so it competes on the only lever it has left, which is cutting price, which shrinks the share further. A factory capturing 70% can afford to invest in the things that let it charge more. The chain you sit in does not just set today's margin; it sets the direction your margin will drift over the next five years.

The hidden costs that do not show up on the invoice

Margin compression is the cost you can see. The costs you cannot see are often larger. Here is a rough comparison of the strategic position you hold under each model.

Notice the last row. Margin is not the only thing an opaque intermediary structure can cost you. When the contract, the money flow, and the information all pass through someone you do not really know, payment risk hides in the gaps. Exporters have lost both goods and payment in deals where a middleman sat between them and a buyer they never verified. If your collection process depends entirely on trusting an intermediary's word, you are carrying a risk you cannot see and cannot price. We cover the operational side of this in export compliance traps and the logistics version in overseas warehouse and logistics traps.

Why smart founders stay trapped anyway

If the math is this clear, why do so many capable exporters keep routing everything through agents? Because the trap is comfortable, and going direct feels risky in ways that are easy to overstate. Naming the real reasons is the first step to getting past them.

The four comfortable lies

• "The agent handles everything, so I am free to focus on production." True, until the agent leaves, raises their cut, or finds a cheaper factory. The convenience is real and the dependency is total.
• "I do not speak the language or know the market, so I cannot sell direct." A decade ago, decisive. Today, buyers research in their own language online, and the gap between you and them is a content and visibility problem, not an unbridgeable cultural wall.
• "Going direct means firing my agents and starting a war." It does not. Going direct means stopping the agent from being your only bridge. You can run a direct channel alongside existing relationships for a long time.
• "Direct selling is for consumer brands like SHEIN, not for my industrial parts business." The consumer playbook does not transfer cleanly, but the principle does. For B2B, direct means owning a verified buyer list and your own credibility, not building a storefront.

Each of these contains a grain of truth, which is why they are sticky. The job is not to pretend the risks are zero. It is to see that the risk of staying trapped, slow erosion of margin and zero market knowledge, is the larger and more certain one.

The agent's incentive is not aligned with yours

Worth stating plainly, because it explains a lot of otherwise puzzling behavior. A good agent is not malicious, but their incentives diverge from yours in three ways. First, they want margin spread, so they have a reason to keep your ex-works price low and the buyer's price high, and to keep you from ever seeing the gap. Second, they want to be irreplaceable, so they have a reason to keep you and the buyer apart, because the day you talk directly, their leverage vanishes. Third, they want low risk for themselves, so when a cheaper factory appears, switching is rational for them and catastrophic for you. None of this requires bad faith. It is simply what the structure rewards. Expecting an intermediary to act against their own incentives is not a strategy.

The China context: why now is different

The middleman model was not a scam invented to cheat factories. For most of the last three decades it was a genuine solution to genuine problems. Understanding why it made sense, and why those reasons have weakened, tells you why the window to go direct is open now in a way it was not before.

What the agent used to solve

An intermediary historically bridged four real gaps for a Chinese factory: language, because the factory could not communicate with a buyer in German or English; discovery, because there was no practical way to find the right distributor across an ocean; trust, because a foreign buyer would not wire money to an unknown plant; and logistics and finance, because export documentation, customs, and payment terms were genuinely hard. For a manufacturer in the 1990s or 2000s, paying an agent to handle all four was rational. The agent earned their cut.

What changed

Each of those four gaps has narrowed sharply. Language is now a content and translation problem that a small bilingual team or partner can solve. Discovery has inverted: buyers now research suppliers online and through AI, so being findable matters more than having a contact who finds them. Trust can be built through earned third-party coverage and a credible online presence, visible to a buyer before they ever pick up the phone. Logistics and finance, while still real, are far more accessible than they were. The export infrastructure has matured even as the export volume has surged. The cross-border e-commerce growth cited at the top of this guide is the macro signature of exactly this shift: more value moving with fewer hops between maker and buyer.

The practical implication is that the agent now solves fewer of your problems than they used to, while still capturing the same or larger share of your margin. That is the definition of a structure that has outlived its rationale. The factories acting on this now are buying a head start; the ones waiting are betting that the old structure will protect them, which is the one thing it was never designed to do.

Real cases: who broke out, and how

The brands that escaped the trap did not do anything mystical. They got closer to the buyer, sometimes the end consumer, sometimes the business purchaser, and refused to give that relationship away. The specific tactics differ; the underlying move is identical.

Anker: owning a commodity category

Founded by a former Google engineer, Anker took commodity charging electronics, a category drowning in faceless white-label sellers, and built a direct brand on Amazon and its own channels. Analysts describe its "shallow sea" strategy of owning fast-growing niche categories like power banks and wireless audio, as covered by EPAM's analysis of Chinese DTC brands. The lesson is not "sell on Amazon." It is that Anker captured the customer data and the brand, so it set prices instead of accepting them. A USB cable from an anonymous seller competes only on price. The same cable under a trusted brand competes on trust, and trust holds a premium.

SHEIN: compressing the chain to zero

Whatever you think of its model, SHEIN compressed the distance between thousands of Chinese garment factories and Western consumers to almost zero. By reading demand off social platforms and feeding it back into a test-and-reorder production system, it reportedly reached around $10B in annual revenue in European and American markets, per a SHEIN case study from AllValue. The factories that plugged into it stopped being anonymous suppliers and became part of a demand engine. The principle: the closer you sit to live demand signal, the more of the chain you can collapse.

A B2B components maker: the unglamorous version

Consider an illustrative scenario that plays out constantly in industrial exporting. A precision components maker in Ningbo sells through a trading company that resells to European machine builders. The factory has no idea which builders use its parts. One year, orders from the trading company drop 30%. The factory assumes the market softened and cuts its own prices to keep the trading company happy. In reality, demand was flat; the trading company had simply shifted volume to a cheaper factory while keeping the same end customers, who never knew the parts had changed.

Now picture the same factory after going direct. It knows the names of the machine builders, the procurement managers, and the application engineers who specify parts. When orders dip, it calls the actual buyer and learns the real reason in an afternoon. It does not cut prices into a phantom downturn. "Direct" for this factory never meant a Shopify store. It meant knowing the named distributors, OEM buyers, and procurement managers who actually place orders, and reaching them itself instead of through an agent's Rolodex. That is the harder, less glamorous version of the same move, and it is the one most Chinese industrial exporters still skip. The full mechanics of building that engine are in our China B2B export playbook and the broader B2B growth engine guide.

The direct-to-buyer model, step by step

Going direct is a sequence, not a switch. You do not flip a lever and wake up with a direct business. You build it in order, because each step depends on the one before it. For a B2B exporter especially, the playbook looks like this.

1. Define the real buyer. Not "Europe." A specific role at a specific kind of company. For example, the purchasing manager at mid-size HVAC distributors in Germany doing 5 to 50 million euros in revenue. Precision here decides everything downstream, because a vague target produces a vague list and vague outreach that converts no one.
2. Build a verified prospect list. Company name, the right contact, role, verified business email, and a reason they would care. This is the asset the agent was holding hostage. Owning it is the whole game. A list of 200 precisely-right buyers beats a list of 5,000 random ones every time.
3. Be findable when they research you. Buyers vet suppliers on Google, and increasingly inside ChatGPT, Gemini, and Perplexity, before they ever reply. If you are invisible there, even a perfect email lands cold. This is where search and AI visibility stop being marketing luxuries and become sales infrastructure.
4. Earn third-party credibility. A buyer trusts a factory more when it appears in trade press, on industry sites, and in the AI answers they are already reading. Earned coverage does the convincing before you do, which matters enormously when the buyer has never visited your plant.
5. Run outreach you control. Your own team sends the messages, books the calls, and keeps the relationship. The contract sits with you, not an intermediary. Only your team knows the product, the pricing flexibility, and how much you can actually concede in a negotiation.
6. Measure, learn, and feed it back. Track which buyers reply, which objections recur, which segments convert, and route every insight back into the product and the next list. This closes the feedback loop the middleman used to break.

Notice what changed across these six steps: the buyer relationship, the thing the agent used to own, now lives inside your company. That is the entire point. Everything else is mechanics.

A 90-day version for a team that is starting cold

If reading six steps makes the whole thing feel like a year-long project, compress it. Here is an illustrative 90-day path for a factory with one or two salespeople and no direct channel today.

• Days 1 to 30: Lock the buyer definition. Pick one country and one buyer role. Commission or build a verified list of 150 to 300 prospects in that exact segment. Audit what a buyer sees today when they search your company name and your product category.
• Days 31 to 60: Fix the findability gaps. Publish a credible English (or local-language) site and a handful of pages that answer the questions buyers actually ask. Begin earning one or two pieces of third-party coverage. Draft and test outreach templates on a small batch.
• Days 61 to 90: Run controlled outreach from your own team to the full list. Book calls. Log every reply and objection. By day 90 you should have a handful of direct conversations that no agent introduced, and a feedback file that is already worth more than the orders.

A worked scenario: the same factory, two futures

Frameworks land better when you can see them run. Here is a single illustrative factory, viewed under both models, to make the abstract sequence concrete. The numbers are illustrative, chosen to show the mechanism, not to describe any real company.

The starting position

Picture a mid-size maker of commercial kitchen equipment in Guangdong. It does roughly 60 million RMB a year in export sales, almost all through two trading companies and one long-standing agent who introduced its largest account a decade ago. Gross margin on these orders runs thin, the founder cannot name a single end customer, and every January the agent renegotiates terms in the agent's favor. The factory is profitable but fragile: lose one trading company and a third of revenue walks out the door overnight, to a destination the founder cannot even see.

Future A: nothing changes

The factory keeps optimizing what it can control, which is production cost. It shaves a few points off the bill of materials, runs the line harder, and squeezes suppliers. None of this reaches the customer, because the agent absorbs any savings as extra spread. A competitor in a neighboring province offers the trading companies a slightly lower price; orders quietly migrate. The founder, blind to the real cause, assumes the market softened and cuts prices again, accelerating the decline. Within a few years the factory is a low-margin contract manufacturer with no relationships of its own and no way to climb back up. This is not a dramatic collapse. It is the slow, common version of the trap closing.

Future B: the factory goes direct, carefully

The founder does not fire the agent. Instead, the factory picks one market the agent does not cover well, say independent restaurant-equipment dealers in the United States, and defines the exact buyer: the purchasing lead at regional dealers doing a specific revenue band. It commissions a verified list of a few hundred such dealers. It publishes a credible English site that answers the questions a dealer actually asks about certification, lead time, and warranty. It earns two pieces of industry coverage so a dealer searching the company name finds substance, not a blank. Then its own small sales team, not a vendor, works the list.

The first month is quiet. By the third, a handful of dealers are in real conversations that no agent introduced. The orders are smaller at first than a trading-company bulk order, but the margin per unit is dramatically higher, and for the first time the founder is hearing directly what the market wants. Two years on, direct accounts are a growing slice of revenue, the agent relationship still exists but no longer dictates terms, and the factory has a feedback loop it never had before. Same factory, same product, radically different position, decided entirely by who owns the bridge to the buyer.

Common mistakes and pitfalls

Most direct-to-buyer attempts do not fail because the strategy is wrong. They fail because of a handful of predictable execution errors. Here are the ones we see most often, and how to avoid each.

Treating "direct" as a website project

A surprising number of exporters believe going direct means building a fancy website and waiting. A site is necessary but not sufficient. Without a defined buyer, a verified list, visibility in the channels buyers use, and a team running outreach, a beautiful site is a billboard in the desert. Build the demand path, not just the destination.

Buying a giant, dirty list

Cheap lists of tens of thousands of "leads" are worse than no list. They are full of dead emails, wrong roles, and companies that will never buy. Your team burns weeks emailing into the void, concludes "direct does not work," and retreats to the agent. A smaller, verified, precisely-targeted list is the asset. Quality of targeting beats quantity of names in every honest comparison.

Outsourcing the actual selling

Some vendors offer to "do the outreach for you," contacting buyers in your name and handing you booked meetings. This recreates the exact trap you are escaping. The vendor now sits between you and the buyer, owns the first conversation, and can disappear with the relationship. The outreach, the conversations, and the contract must live with your own team. A partner's job is to hand you the verified list and the visibility, then step back.

Going dark in search and AI

You can have a perfect list and a strong product and still lose, because the moment a buyer receives your first email, they search your company name and your category. If the results are empty, or worse, dominated by competitors, your credibility collapses before the conversation starts. Visibility is not a separate marketing task; it is the silent half of every outreach. The shift from classic search to AI answers makes this more urgent, which we unpack in GEO versus SEO in 2026 and how to get cited by AI.

Quitting at the first quiet month

Direct relationships compound, but slowly at first. The first month of outreach is almost always quiet. Founders accustomed to the agent's steady purchase orders panic and abandon the effort right before it would have paid off. Set the expectation up front: months one and two build the pipe, months three onward fill it.

Metrics to watch

You cannot manage what you do not measure, and the middleman model trained you to measure almost nothing about the market. As you go direct, track a small, honest set of numbers. Resist the urge to track everything; these are the ones that tell you whether the bridge is yours.

If you track only one of these, track direct revenue share. It is the single number that proves the bridge is moving into your hands. Everything else is a leading indicator of it.

A quarterly review you can actually hold

Metrics only help if someone looks at them on a rhythm. Set a simple quarterly review with three questions. First, did direct revenue share rise, and if not, where did the pipeline stall: list, visibility, or outreach? Second, what did the market tell us this quarter that we would never have heard through an agent, and did that insight reach the product team? Third, are we more or less concentrated in one buyer than three months ago? Three honest answers per quarter keep the effort from drifting back into the comfortable old habit. The discipline is not in the dashboard; it is in the meeting where someone has to explain the numbers out loud.

Choosing a partner without recreating the trap

The cruel irony of going direct is that the market is full of vendors who promise to help and then install themselves as the new middleman. You escape an agent only to acquire a "growth partner" who owns your buyer relationships, your accounts, and your data. The structure repeats; only the name on it changes. Here is how to tell a genuine enabler from a fresh trap.

Questions that expose a hidden middleman

• Who sends the outreach? If the answer is "we do, in your name," walk away. A real partner hands you the list and the templates; your team sends. The moment a vendor controls the first conversation, they control the relationship.
• Who owns the data and the accounts? The verified list, the contacts, the replies, and the contracts must be yours, portable, and exportable. If leaving the vendor means losing your buyers, they are a middleman with a dashboard.
• How are they paid? A fee for defined work (a list, a visibility program, earned coverage) keeps incentives clean. A cut of your sales gives them a permanent claim on your customer relationship, which is the exact thing you are trying to own.
• What happens if you stop working with them? A good partner leaves you stronger and self-sufficient: you keep the list, the rankings, the coverage, and the playbook. A trap leaves you dependent, with the assets walking out the door alongside the vendor.

Apply these four questions to every vendor, including us. The point of going direct is sovereignty over your buyer relationships. Any arrangement that quietly transfers that sovereignty to a third party, however modern its branding, is the same old trap in new clothes.

The build-versus-buy decision

You will face a choice on each piece of the playbook: build the capability in-house or commission it from a partner. A reasonable default is to own the parts that are your relationship (the outreach, the conversations, the contracts) and to buy the parts that are specialist craft and slow to build internally (a verified list at quality, search and AI visibility, earned media). Building a credible visibility presence from zero can take a small team many months of trial and error; buying it from people who do it daily compresses that. But never outsource the relationship itself. The dividing line is simple: outsource the infrastructure, own the customer.

How Ignite helps, and where the line is

This is the work we do for Chinese exporters going global. We want to be precise about what that means, because the category is full of vendors who blur it, and the blur is exactly how new middlemen are born.

On the lead-generation side, we deliver a researched, verified prospect list of your ideal overseas buyers, as a clean Excel or CSV file, plus free outreach templates your team can adapt. We do not cold-contact buyers on your behalf or pretend to be you. Your sales team runs the outreach and owns every relationship from the first reply. That distinction is the whole reason the model works: the buyer becomes your customer, not ours. We never sit between you and the people you are trying to reach.

On visibility, our China market programs and SEO/GEO work make sure that when a buyer Googles your company or asks an AI which suppliers to consider, you actually show up, credible and in their language. On credibility, our digital PR earns the third-party coverage that makes a serious buyer believe a factory they have never visited; we go deep on why that coverage is a durable moat in digital PR as the GEO moat. For influencer and KOL programs, we charge an agency fee for the work, not a cut of your sales, and creator fees are always separate and transparent.

None of this requires you to fire your existing agents tomorrow. It requires you to stop letting them be the only bridge to your market. The exporters who thrive over the next decade will be the ones who own that bridge themselves.

Frequently asked questions

Is using a middleman always a mistake?

No. A good distributor that holds inventory, extends local credit, handles after-sales service, and genuinely opens a market earns its margin. The trap is not having a partner; it is letting that partner be your only link to the market, so that you own no relationship, no data, and no pricing power of your own. The goal is to add a direct channel you control, not necessarily to remove every intermediary.

How is "going direct" different for B2B versus consumer products?

For consumer products, direct often means a branded storefront or marketplace presence where you own the customer data, as Anker and SHEIN do. For B2B, it almost never means a storefront. It means owning a verified list of the specific companies and roles that buy your category, being findable when they research, and having your own team run the outreach and the relationship. The principle (own the buyer relationship) is identical; the tactics differ. We contrast the channel choices in DTC versus platforms cost truth and the B2C growth playbook.

Will I damage my relationship with my current agents if I start selling direct?

Not if you sequence it sensibly. Many exporters run a direct channel in new markets or new segments where the agent is not active, leaving existing arrangements untouched at first. The point is to reduce single-bridge dependency, not to declare war. Over time, as your direct channel proves itself, your negotiating position with every partner improves, because you are no longer captive.

How much should a verified buyer list cost, and how big should it be?

Pricing varies by how narrow and how thoroughly verified the segment is. The more useful question is size: a tight list of a few hundred precisely-right buyers in one country and one role is worth far more than thousands of loosely-matched names. A small, accurate list your team can actually work through in a quarter beats a giant one nobody touches. Always favor verification and targeting over raw volume.

Why does search and AI visibility matter if I am emailing buyers directly?

Because the email is only the opening. The moment a serious buyer is interested, they search your company name and your product category, and increasingly they ask an AI tool for a shortlist of suppliers. If you are invisible or your competitors dominate those results, your credibility evaporates before the first call. Visibility is the silent half of direct outreach. See how Google AI Overviews change buyer research for what this looks like in practice.

How long before going direct pays off?

Expect the first month or two to be quiet while you build the list, fix visibility, and start outreach. Direct relationships compound: the pipeline you build in months one and two tends to fill from month three onward. Founders who quit during the quiet phase never see the payoff. Treat it as infrastructure, not a campaign.

Does Ignite contact my buyers for me?

No, and this is deliberate. We deliver the verified list and the visibility, then step back. Your team sends every message, has every conversation, and signs every contract. If a vendor offers to run outreach in your name, understand that they are recreating the middleman trap with themselves in the middle seat. The whole value of going direct is that the relationship is yours.

What if I sell a true commodity with no brand potential?

Even commodities benefit from owning the buyer relationship and the data. You may not command a brand premium on a generic part, but if you know your buyers directly, you keep the distributor margin, you hear about demand shifts first, and you are not blind to why orders move. Anker proved that even "commodity" charging electronics can carry a trust premium once a brand owns the customer. The floor case for going direct is captured margin and market sight; the ceiling case is brand pricing power.