Planning the structure of an export site: the three decisions that come first

A client came to us with an electric bike range sold across several Southeast Asian countries. There were no page requirements — no wireframes, no reference sites, no list of sections. That sounds like freedom and it is mostly a burden, because it means every structural decision is yours and you will be the one who is wrong if it is wrong.

We made three decisions before writing a word of copy. They are the same three on every export site we have built since.

One: how the range is organised

A buyer arrives knowing one of three things — the model name, the use case, or nothing at all. The catalogue has to work for all three, and the organising axis you pick decides which of them you serve well.

Organising by internal series numbers is the most common mistake, because it is how the company thinks about its own products. The buyer does not have that map. They have a problem, and the categories have to be named after that problem.

Two: what a product page must answer

This is the decision that costs the most to get wrong, because changing it later means editing every product. Write the list once, then every product page has the same slots in the same order — and a missing answer becomes visible instead of being quietly absent.

• What is it, in one line a non-specialist understands.
• The specification, as a table. Not prose. A buyer comparing three models reads three tables, and prose makes that impossible.
• What is in the box, and what is not.
• What it costs, or if the price is not public, exactly what happens when they ask.
• How long it takes to reach them, and who handles customs.
• What happens when it breaks — warranty term, and who they contact.

The last two are the ones that get left off, and they are the ones a serious buyer needs most. Someone comparing suppliers from another country is not choosing a product, they are choosing a risk. Delivery and after-sales are where that risk lives.

Three: where the purchase path begins

On this site the answer was a full WooCommerce checkout, because the bikes have list prices and are sold to individuals as well as dealers. That is not the right answer everywhere. If your prices are negotiated, a cart is a liability — it puts a number on the page you did not intend to commit to.

Mixing them is the failure case. A page with both a cart button and a "request a quote" button asks the buyer to decide which kind of customer they are, and they will pick wrong or pick neither.

Why structure first, and not content first

Content written before the structure exists has nowhere to go, so it becomes paragraphs. Paragraphs cannot be compared, cannot be filtered, and cannot be read by a machine that is trying to answer a question on your behalf. Structure first means the writing has slots to fill, and an empty slot is a question you have not answered yet — which is exactly the thing you want to see before a buyer sees it.

The site runs on standard WordPress and WooCommerce. That was deliberate: the client can move it to any host and hire anyone to change it. A structure decision that locks you into one agency is not a structure decision, it is a hostage situation.