Dropshipping in 2026: the €3 duty changed the arithmetic, and multichannel is what still works

Dropshipping is living proof that one business model can be both an overhyped route to “fast money online” and a real strategy for thousands of sellers who earn from it predictably, month after month. The difference between those two groups isn’t luck. It’s numbers, suppliers and automation.

This post is for the second group. We sell nothing to dropshippers, so we can be blunt about it: here is whether the model still adds up in 2026, what actually changed, and how to run several channels without losing control of them.

What changed on 1 July 2026

If you have spent the last few years dropshipping goods shipped directly from China to EU customers, the arithmetic stopped working in July.

On 1 July 2026 the EU removed the de minimis customs exemption and introduced a flat €3 duty, charged per item on non-EU consignments worth up to €150 and independently of any VAT due. It’s explicitly transitional: the €150 threshold is expected to be removed altogether around 2028.

Run it against your own order. If a sale carries a few euros of margin, €3 per item is not a squeeze — it’s the entire profit. Add delivery measured in weeks, in markets where buyers now expect next-day, and the model loses on price and on service at the same time.

The model that still works: EU warehouses and honest margins

Dropshipping from suppliers with warehouses inside the EU is a different business. Delivery lands in 24–48 hours, returns drop, ratings hold — and the margin survives contact with reality.

Before counting revenue, count what’s left after costs:

• Gross margin: typically 20–35% of the sale price, depending on category and your relationship with the supplier.
• Minus marketplace commission: on Allegro, 1% to 17% net depending on category — plus loyalty-programme surcharges and any advertising spend. Other marketplaces differ, but none of them are free.
• Minus payment costs, returns, tooling and support.

What survives is usually a single-digit to low-double-digit percentage of revenue. The workable band is roughly 10–25% margin after every cost — below that, fixed costs eat the profit no matter how much volume you push through.

That last sentence is the one worth sitting with, because it decides which tools you can afford. In a business running on thin per-order margin, a platform priced per order is taking a cut of the only number that matters.

One channel is a single point of failure

Dropshipping on one marketplace is simpler, and that is exactly what makes it fragile. Commissions, ranking algorithms and category rules change regularly and without long transition periods. When your only channel changes the rules, your business changes with it.

Sellers who are still earning from dropshipping in 2026 do what every durable e-commerce operation does: they diversify. A marketplace, a second marketplace — in Poland that pairing is usually Allegro and ERLI — and their own store on WooCommerce or PrestaShop. More buyers, and no single algorithm change that can end the business.

The objection is always the same, and it is a fair one: doesn’t that triple the operational work?

Three channels, three panels — the actual problem

In theory, three channels. In practice, three admin panels, three sets of stock levels and three places an order can appear. That is why plenty of sellers stay on one platform — and why the ones who do go looking for a fix tend to bounce off SaaS channel managers whose bill grows with their sales.

The mechanics are simple. When the last unit sells on one marketplace, stock has to drop on the other channels immediately. A slow-moving catalogue can live with polling every one to three hours. Dropshipping at volume cannot: the safest setups combine event webhooks with periodic polling as a backstop, which we covered in how order and inventory sync actually works.

Manual management falls apart somewhere around 20–30 orders a day. Past that you start selling stock you don’t have, shipping the wrong things, and collecting the negative reviews that — in a business with no brand of its own — are the one asset you cannot buy back.

The four layers dropshipping automation needs

Dropshipping automation is not one plugin. It is four layers:

1. Inventory sync — supplier availability reaching every channel at once. Without it you’re selling what you don’t have.
2. Order routing to the supplier — the marketplace order reaches the wholesaler without you touching it; you get a tracking number back.
3. Price sync — when the supplier’s purchase price moves, your sale price recalculates against your margin rule. Without it you drift below profitability and find out later.
4. Invoicing — a document raised per order, automatically.

Layers 1 and 4 are solved problems. Layers 2 and 3 depend entirely on whether your supplier exposes an API at all — which is worth asking them before signing, not after.

Where OpenLinker fits, and where it does not

If you run a WooCommerce or PrestaShop store and sell on marketplaces alongside it, syncing stock, orders and invoices is your daily operational problem.

We are building OpenLinker — an open-source (Apache 2.0), self-hosted platform that connects those channels in one place. It runs on your own server, so order and customer data never lands in someone else’s cloud, and there is no per-order fee — you pay for infrastructure, not for every unit you sell. Against the margins above, that is not a cosmetic difference.

Let us be precise about what is live today — this is a post about numbers, so our own claims should be checkable too. Shops PrestaShop and WooCommerce; marketplaces Allegro and ERLI; shipping InPost and DPD; invoicing Subiekt nexo, KSeF and inFakt. That covers layers 1 and 4 above — cross-channel inventory and order sync, and automatic documents.

What you don’t get out of the box is layers 2 and 3: there are no ready-made wholesaler connectors. Integrating a specific supplier means writing an adapter today — which is precisely what the plugin architecture exists for, and why an agency ends up writing the adapter and not the plumbing underneath it. OpenLinker is alpha, pre-1.0; we keep the integrations scorecard current so nobody confuses “planned” with “working.”

Where to go next

• See what’s live today — the honest scorecard, gaps included.
• Try the live demo — the OpenLinker admin on sample data, nothing to install.
• Read the architecture — how a supplier adapter gets written against a stable core.
• Compare against BaseLinker — where “no per-order fee” comes from, and when it actually pays off.
• Star the repo — if you want this to exist sooner.

Dropshipping still pays in 2026. It just stopped paying the people looking for a shortcut.