Cross-Border Commerce Is Broken – $149M Says This Is How You Fix It

01 /The Concept

This startup was founded in 2022, but flew almost completely under the radar for its first few years. In spring 2024, it quietly raised $9M. A year later, in spring 2025, another $40M – still largely unnoticed. Then a few days ago it raised $100M, and people suddenly started paying attention.

The company has just 600 client businesses – which, against $149M in total funding, sounds surprisingly thin. So what exactly is it doing?

Swap's core promise is "sell internationally with confidence." Shipping cost, cross-border taxes, and import duties are calculated automatically and bundled into the price the international customer sees at checkout. Swap also pre-files all the necessary customs and tax compliance documentation.

Payment goes directly to the seller, not through Swap as an intermediary. But Swap can add locally preferred payment methods to the checkout form – including regional wallets and alternative payment instruments common in the buyer's country.

Delivery is managed by Swap as well. Sellers can use their existing carriers or Swap's logistics partners. Either way, duties and taxes are already accounted for and the paperwork is pre-filed – so there should be no customs delays.

Returns are handled just as smoothly, even for goods sold internationally. Swap manages that logistics chain too. This matters: e-commerce return rates average around 20%

Swap even includes a module that surfaces optimization recommendations – from choosing the right logistics partners to adjusting pricing by market.

Access to international sales lifts merchant revenue by an average of 8%. And Swap's full service stack improves margins by an average of 30% compared to merchants trying to manage cross-border complexity themselves. As a result, 98% of merchants who start using Swap stay.

02 /Why It Matters

Cross-border e-commerce is a massive market, projected to approach $1 trillion by 2032.

But most small online merchants ignore this opportunity – because the hassle of managing international sales feels like more trouble than it's worth. Several startups moved into this space years ago, helping local merchants go global by absorbing all the cross-border complexity themselves.

OpenBorder ([related review](/review/pljus-35-vyruchki-za-150-minut)) claims that cross-border sales are viable for any merchant and that even without additional marketing, simply enabling international purchases can grow revenue by 2–6%. With active effort behind it, 10–20% or more is achievable for products with global appeal. OpenBorder has raised $10M.

European competitor eBrands ([covered here](/review/tema-v-kotoroj-deneg-bolshe-chem-zhelajushhih-ej-zanjatsja)) has raised nearly €50M, including €7.5M last spring.

But the most interesting thing about Swap isn't the cross-border logistics layer – it's the conceptual bet Swap is making about the future architecture of e-commerce, which it calls "Agentic Commerce." The word "agentic" here means exactly what it sounds like: commerce run by AI agents capable of reasoning, planning, and managing the full sales process from product discovery to delivery.

The key distinction is that AI agents can operate across every customer touchpoint and integrate with any external platform.

This is a more radical claim than it first appears – and one that not everyone has fully absorbed yet. One publication described Swap as "a Shopify competitor." That's not quite right, or maybe not right at all in the conventional sense.

Shopify is listed among Swap's possible integrations – alongside cloud services, shipping providers, and other e-commerce tools.

Swap is a layer of AI agents that integrates with other platforms and services. It sits either beneath them as an orchestrating foundation or above them as a conductor managing the entire ensemble. What it is not is a direct equivalent of Shopify – where you "merely" create a store.

Swap pushes this framing further – announcing that "the.com era is over" The idea being that old-school e-commerce was defined by owning a domain, driving traffic there, and transacting on that URL.

But that model is already eroding. Active marketplace sellers often don't need their own websites at all. Many sellers in Asia conduct their entire business inside WeChat or WhatsApp.

Swap's vision of Agentic Commerce is "a fundamentally new way to sell" – AI agents embedded everywhere: every online and offline channel, any customer touchpoint that exists. But those agents share a single product catalog, unified pricing logic, and a common operational base. The result is seamless selling from any channel, with full visibility into the whole operation.

The center of gravity shifts: instead of a website with awkward module integrations bolted on for every additional sales channel, the platform itself becomes the anchor. The website, if it even exists, becomes just one more channel plugged into the same system.

03 /Opportunities

The most consequential competition rarely happens between direct substitutes. It happens when a fundamentally new product makes an entire category of existing products obsolete.

Henry Ford observed that customers in his era wanted "a faster horse" – but what he gave them was the automobile, which replaced horses as a category. ChatGPT became Google Search's most significant competitive threat not because it's a search engine, but because it answers questions without requiring you to search at all. Neither of those disruptions looked like conventional competition until the damage was already done.

The opportunity this points toward: stop designing better versions of existing products. Instead, think hard about what entirely different product could eliminate the need for a whole category of current solutions.

What would that product look like in your space? How would it work? Which legacy category would it make unnecessary?

If no obvious answer emerges, "Agentic Commerce" platforms – which could eventually replace both website builders and marketplace storefronts for the right kind of merchant – offer a concrete starting point for that kind of thinking, with Swap as the reference case.