Landed cost is the new tiebreak: how cross-border merchants earn AI citations for their all-in price

Cross-border merchants notice a strange pattern: their product is cheaper on the storefront than any local competitor, yet an AI shopping assistant keeps recommending the local option. The assistant is not ignoring you. It is quietly comparing landed cost - the all-in number that includes the base price, shipping, duties and taxes, and delivery time - and you never gave it the data to build an accurate estimate for your product.

When an assistant cannot resolve your true delivered price, it does not stop answering. It estimates from whatever signals it can find: your shipping policy page, vague "varies by destination" lines, or the absence of any duties information. Faced with uncertainty, it downgrades you and recommends the merchant whose all-in cost it can actually state. This article explains how assistants assemble landed cost, why hidden fees quietly kill citations, and how to publish the machine-readable facts that win the comparison.

An AI assistant answering "which of these ships fastest and costs least to Berlin?" does not run a checkout simulation. It strings together facts it can read and verify:

• Base price from the visible price and its Offer structured data, including currency.
• Shipping cost and threshold, read from the shipping policy or a shipping-rate table in text and structured data.
• Duties and taxes, which are the least-published fact and therefore the biggest source of uncertainty.
• Delivery estimate, often the deciding factor between two otherwise-equal offers.

Many assistants do not just read currency; they convert it at query time and reason about whether customs and VAT are included or added later. If your INR or CNY price is stored without a clear country and currency scope, a German buyer's session sees an offer it cannot fully trust.

The most reliable way for a cross-border merchant to be named is to remove estimation from the equation. Three concrete moves consistently raise citation share:

1. Publish a clear delivery and duties statement per destination market. A short sentence frame - "Price excludes import VAT. Duties calculated at checkout; delivery to the EU in 4-7 business days, duties paid" - gives the assistant a citable, low-risk fact. Put the same facts in your ShippingDetails and Offer shipping markup. 2. Model "duties paid" (DDP) where you can. When you handle duties at checkout, you can state an all-in number confidently. Assistants favor the merchant that states a single, complete delivered price over one that leaves VAT to be discovered later. 3. Make the numbers visible as text, not only inside checkout. If duties and delivery estimates live only behind a login or inside a JavaScript-populated cart, retrieval cannot read them. Keep a readable, crawlable summary on the product or shipping page.

The gap between "we ship everywhere" and "here is exactly what it costs to land in Frankfurt in 5 days" is where citations are won or lost. A free scan on geovisora.com/audit reads a product URL the way an assistant does and flags exactly which price, shipping, and duties facts are readable versus buried in checkout-only or image-only places - so you can close those gaps instead of guessing.

Beyond prose, three schema blocks let an assistant cite exact numbers instead of estimating:

• Offer with a price, a priceCurrency, and an availability kept in sync with what humans see; a stale price is the fastest way to lose an all-in comparison.
• ShippingDetails with a shipping rate and delivery-time ranges per region.
• MerchantReturnPolicy and any duties or import notes, so the assistant can state your return and customs posture rather than assume.

Keep these mirrors of your visible copy. When text and structured data disagree, retrieval treats the page as inconsistent and drops it from the confident-answer tier.

International buyers ask in their own language and compare in their own currency. A store that states "1,299 EUR delivered, duties included" on a German-market page is dramatically easier for a German-language session to cite than one whose English product page forces the assistant to add shipping and guess import VAT. Localize the offer fields and the delivery statement together - not just the marketing copy - for each market you actually fulfill into.

_Is it enough to lower my storefront price and absorb shipping quietly?_ It helps the human, but not the assistant, because the assistant still needs a single stated delivered figure. Be explicit about "shipping included" and any duties so the number it reads is the number it can quote.

_Does DDP cost me the price advantage?_ Often the opposite. When you publish duties-paid pricing, you control the number the assistant quotes. Merchants who hide duties let the assistant estimate a higher, more pessimistic total - and the competitor that states the complete price wins the cite.

_My catalog has hundreds of SKUs and many markets. Where do I start?_ Start with your top hero SKUs and your two largest export markets. Publish a per-market delivery and duties statement for those, mirror them in structured data, then expand. Audit the current state first with a tool that shows exactly which price, shipping, and duties facts are currently readable, so the fix targets the real gaps.