Designing an Online Store: From Platform Choice to Conversion Rate

A company site that looks good and lists the right phone number has largely done its job. An online store is not like that. Here every detail — the order of the form fields, the number of clicks to payment, the way stock is shown — lands directly in the sales figures. That is why store design cannot be treated as a graphics project. You are building a selling system.

What follows is the order that makes a difference in real projects: the structural decisions first, then the data model, then the buying experience, and measurement last.

1. Off-the-shelf platform or custom build?

If the catalogue is small, the sales process is standard and there is no unusual logic to implement, an off-the-shelf platform is almost always the better choice: it launches sooner, costs less, and you are not re-solving problems somebody already solved. But if your pricing depends on the customer or the quantity, if the product is configurable (picking dimensions, fabric and fittings in the furniture trade, for example), or if orders have to flow into production and inventory systems, a ready-made platform turns into a ceiling very quickly.

2. Take the product data model seriously

This is the least glamorous and most expensive decision in the project. A wrong product structure cannot be rescued later by good design. Before anything starts, it should be clear how variation (colour, size, material) is modelled, whether stock is held at product level or per variant, and where the price comes from.

• How many variants does each product have, and which combinations actually exist?
• Is the price fixed, or does it depend on quantity, customer and date?
• Which technical attributes exist, and which of them need to be filterable?
• How many levels does the category tree have, and can one product sit in several categories?
• Do images belong to the product or to each variant?

The answers to those five questions determine the product page, the category filters, the cart, and even the URL structure. In our experience, teams that skip this step end up doing a data migration six months later.

3. Keep the path to purchase short and unambiguous

Most of the drop-off in online stores happens at two points: the product page and checkout. On the product page, what the buyer is looking for is usually not complicated — the final price, whether it is in stock, when it arrives, and the returns terms. Leave any one of those four vague and the buyer either calls to check or closes the page.

At checkout, every extra field has a cost. Forcing people to register before they can buy is one of the most expensive decisions available to you; allow guest checkout and offer an account after a successful payment. Show the shipping cost before the final step as well: a charge that appears out of nowhere at the end is the single biggest reason carts get abandoned.

Every question a buyer has to phone you to answer is one lost order and one extra call for the support team.

4. The realities of payment and shipping in Iran

A payment gateway is not just a button. Decisions have to be made: what happens to an incomplete transaction, what process runs when money leaves the customer's account but no order is recorded, and how refunds are handled. Shipping is the same: what the rate is calculated from (weight, destination city, cart value), and what becomes of a multi-item order when part of it is out of stock. In Iran both layers carry local specifics — payments run through licensed payment service providers tied to the domestic banking network, and delivery mixes the national post with private couriers whose coverage varies city by city. These are operational decisions rather than technical ones, but if the design ignores them, support ends up carrying the weight.

5. Store SEO: where most of the opportunity is wasted

Stores usually have hundreds of product pages, and that is their largest SEO asset — provided it is managed properly. Three common mistakes: variant pages that produce duplicate content with no canonical, category filters that generate an unlimited number of crawlable URLs, and product descriptions copied from the supplier's site.

1. Write a unique title and description for every product page; even two original sentences beat copied text.
2. Index or block filter pages by a clear rule, not by accident.
3. Add Product structured data with the real price and stock status.
4. Decide what happens to a discontinued product: redirect to a replacement, or an "unavailable" page with similar suggestions.
5. Do not forget internal links from articles and category pages into products.

6. What to measure after launch

For a store, visit counts mean almost nothing. The numbers that actually drive decisions are conversion rate split by device (look at mobile and desktop separately; the gap is usually large), cart abandonment and the step where people drop out, average order value, and the share of returning customers. Without those, every change you make to the site is a guess.

In closing

A successful online store is not the product of striking visual design; it is the product of good decisions in the data layer, the path to purchase and day-to-day operations. Get the product structure right, keep the path short and the measures clear, and the site stops being a catalogue and becomes a sales channel you can improve month by month.