Ecommerce Development
Platform First, Then The Store

A good fit if…

• You are already selling — through messages, a marketplace, a spreadsheet or in person — and the admin around it has quietly become the job.
• Your catalogue has real structure: sizes, colours, weights, pack quantities, bundles, or the same item sold three different ways.
• Stock counts have to stay honest across every channel you sell through.
• You take money from buyers in more than one country, or expect to inside a year.
• Someone on your side will run the store daily, and that person is not a developer and should not have to become one.

Not yet, if…

• You have four products and sell most of them face to face. A website with a clear enquiry path costs less to build and far less to keep alive.
• Nobody can find you. A new store does not fix distribution — web design with SEO built in, or paid traffic, is the earlier spend.
• Product photography and descriptions do not exist yet. That is the long pole on every store build, and part of it is a positioning problem worth solving in branding and identity first.

Shopify

Shopify is the sensible default when you want the platform to carry the parts that are expensive to get wrong — hosted checkout, PCI scope, fraud screening, stored card updates, staying upright during a sale. You pay for that in a monthly fee and a slice of each order, and you accept that the checkout behaves the way Shopify has decided it behaves. For a conventional catalogue with conventional delivery and a team that would rather sell than maintain software, that trade is usually the right one.

It stops being the right one when your pricing logic is unusual, when the store is secondary to a large body of published content, when different customers must see different catalogues, or when the fee bill has grown past the cost of running the alternative.

WooCommerce

WooCommerce earns its place when the store sits inside a site that also has to publish — guides, specifications, recipes, case studies, a real editorial habit — or when the pricing, delivery and fulfilment rules are yours rather than the industry standard. You own the hosting, the updates and the security posture, which is a genuine recurring cost rather than a footnote. In exchange, nothing about the checkout is off limits.

It is the wrong choice for a team with nobody to keep plugins current. An unmaintained store is a slower disaster than a restrictive one, and it always arrives on the day traffic is highest.

A custom build

Custom is for the cases where both of the above spend their lives being fought: account level price lists, quotes that convert into orders, subscription logic tied to your own billing, or an inventory system elsewhere that is the real source of truth. It is quoted per project, and we will usually try to talk you out of it first, because the cheapest store to run is the one somebody else keeps patched.

How we settle it

Four questions, asked before any design: how many products and how do they vary; what payment methods do your buyers already trust; what other systems must the store speak to; and who edits it on a Tuesday afternoon. Those answers point at one platform far more often than they leave a genuine tie, and we write the reasoning down so the decision can be reviewed later rather than re-argued.

Display currency and settlement currency are separate decisions

What a buyer sees and what lands in your account do not have to match. A store can quote in AUD, INR, GBP or USD while settling into one account, and the question worth deciding before the catalogue is priced is whether converted figures get rounded to retail-looking numbers or left as raw conversions that change with the rate. Rounded is friendlier and needs periodic review; raw is accurate and looks like an accident.

The expected methods differ by market

Card acceptance is table stakes everywhere and settles nothing on its own. Buyers in India reach for UPI before anything else, with cards and netbanking behind it. Buyers in the United Kingdom expect cards, PayPal and increasingly a pay-later option at checkout. Buyers in the United States expect cards and the phone wallets they already have set up. Buyers in Australia expect cards and instalment providers. A store selling into more than one of those needs its gateway chosen with that list in hand, not discovered afterwards.

Delivery, duty and the fine print at the border

Cross-border orders fail late and expensively. The two common causes are a delivery price that was a guess and a duty or import charge the buyer never agreed to. We encode the rules your accountant and your courier give us, show the total before the payment step, and put the terms where a hesitant buyer will actually read them. What we do not do is give tax advice — the numbers come from your side, and the store enforces them consistently.

Refunds, chargebacks and the unglamorous half

A refund path that works in the admin, a returns policy written in your own words, and order records complete enough to answer a chargeback are worth more than another promotional banner. Most stores discover this in month three; it is cheaper to build it in month one.

Weeks one and two

Catalogue mapping, platform decision, and the design of the product page, cart and checkout. This is also when product data collection starts on your side, because it takes longer than anyone estimates.

Weeks three to five

The build itself: templates, catalogue import, delivery rules, gateway integration and the admin. You get a staging link partway through rather than a reveal at the end.

Weeks six to eight

Testing the paths that fail, loading final content, wiring analytics and conversion tracking, redirect mapping if a store already exists, then launch and handover. Stores replacing an existing one sit at the longer end; stores starting clean with data ready sit at the shorter one.

What actually causes delay

Product descriptions that were going to be written last week. Photography reshoots. Variant structures that change after the import. A merchant account still in review. None of these are development problems, and all of them are solvable weeks earlier than they usually are.