Starting an online store is an operations launch, a working store needs a validated offer, real payment and shipping setup, and a tested order, not just a homepage.

A store can look completely finished (clean theme, good photography, a working "buy now" button) and still fail the moment a real customer places a real order, because nobody tested whether the payment actually settles, the courier actually collects, or a confirmation email actually arrives. Design is the visible ten percent. Operations is the ninety percent nobody sees until it breaks.

This guide sequences a launch in the order that avoids expensive rework: validate what you are selling and at what real margin, pick a platform with South African constraints already factored in, configure payments and shipping honestly, prepare the catalogue properly, and run one genuine test order before you tell a single customer the store is open.

Treat the storefront as the thin, visible layer sitting on top of a system that has to actually work underneath, a payment gateway, a courier account, stock tracking, and a person who replies when something inevitably goes wrong in the first month.

Add up every cost that will actually apply once the store is live: product or landed cost, packaging, the payment gateway’s percentage plus fixed fee, courier cost, and a realistic allowance for returns or damaged-in-transit stock. A product that looks comfortably profitable on cost price alone can turn out thin, or genuinely loss-making, once all of that is included.

If you are not sure demand is real yet, a cheap test (a landing page with a genuine waitlist, or a small pre-order run) costs far less than building a full catalogue for a product nobody actually wants at the price you would need to charge to make margin work.

Decide your minimum viable catalogue deliberately. Fifteen well-photographed, well-described products with accurate stock usually convert better, and are far easier to manage properly, than two hundred thin listings you cannot yet support with real data or inventory.

Shopify and WooCommerce remain the two most common choices for South African stores. Shopify gives you a more polished checkout out of the box and easier day-to-day maintenance; WooCommerce gives you more flexibility and suits stores with unusual product logic, at the cost of more hands-on management.

The constraint every South African store has to plan around from day one: Shopify Payments, Shopify’s own built-in processor, does not support South African merchant accounts. Every Shopify store here has to connect a third-party gateway instead, and that gateway’s fees are a real, ongoing cost that most generic "how to start a Shopify store" guides never mention, because they were written for a market where the native processor already exists.

None of this is a reason to avoid Shopify, it remains a genuinely strong platform for South African merchants. It is a reason to model gateway fees into your pricing from the very first spreadsheet rather than discovering them after your first month of live sales. See our dedicated breakdown of Shopify transaction fees in South Africa for the specifics.

Sizing charts, material or ingredient lists, and honest photography of colour and scale prevent disputes and returns later, this matters more than how persuasive the description sounds on first read.

Write specific descriptions for every product instead of short generic labels. "100% cotton, true to size" answers a real question a buyer has; "premium quality" answers nothing. A buyer cannot touch the product before paying, your data and photography are the only substitute for that missing sense of touch.

Connect stock levels to your actual inventory count wherever the platform allows it. Overselling something you do not actually have on hand creates a support and trust problem that costs far more than the original sale was ever worth.

South African shipping options typically include courier-to-door, PUDO or Paxi-style locker and retail collection points, and in some cases your own limited-radius local delivery. Each carries different cost and speed trade-offs worth testing before you commit to one as the default at checkout.

State delivery windows honestly, and separately, for major centres versus more remote areas, a single blanket "3 to 5 working days" claim that is only true for Johannesburg and Cape Town creates avoidable, entirely predictable complaints from buyers everywhere else in the country.

Decide returns logistics before launch, not after the first return request lands in your inbox: who pays return shipping, what condition items must arrive in, and how refunds are processed and communicated back to the buyer.

• A returns and exchange policy written in plain language, consistent with Consumer Protection Act cooling-off requirements for online sales.
• A privacy notice covering what customer data you collect and how it is used, in line with POPIA.
• Clear terms of sale, including pricing, delivery timeframes and how disputes are handled.
• A decision on VAT registration if turnover is approaching or above the compulsory registration threshold, confirm current thresholds with SARS or your accountant.
• Business or sole-proprietor registration appropriate to how you intend to trade and invoice.

1. Place a genuine order as a customer wouldUse a real card or gateway transaction in test or live mode, not just an internal admin-panel order.
2. Confirm the payment actually settlesCheck the gateway dashboard reflects the transaction correctly, including any fee deduction.
3. Fulfil and ship it for realPack, label and dispatch through your chosen courier exactly as a live order would be handled.
4. Check every automated messageOrder confirmation, shipping notification and any follow-up email should arrive correctly and read clearly.
5. Review the full experience as the customerFrom ad or link click through to receiving the product, note any point that felt unclear or slow.

The store build itself is one line item among several ongoing costs, not the whole budget. Nexus ecommerce builds start from R7,500 for a properly configured Shopify or WooCommerce store, platform subscription fees, payment gateway charges, courier costs and any paid apps are separate, ongoing costs that continue every month after launch, not once-off items.

Budget for all of that as a monthly operating line, not a single upfront number. A store that is "built" but has no plan for its recurring platform, gateway and courier costs will look profitable on the sales dashboard while quietly losing money once every real cost is actually included.

If budget is tight at launch, prioritise spend on what a customer directly experiences and what protects trust (accurate product data, a tested checkout, honest delivery timeframes) ahead of cosmetic extras that make no difference to whether an order actually completes.

In the first two weeks, watch every single order manually: confirm payments settle correctly, fulfilment happens on schedule, and support messages get a same-day reply. Small stores can and should be hands-on at this stage, problems caught early cost little; the same problems discovered after a paid traffic push cost a lot more.

By day 30, review your actual numbers: conversion rate, average order value, real margin after gateway and courier fees, and any recurring support question that reveals a gap in your product data or policies. Fix the gap before adding more traffic on top of it.

By day 60, decide deliberately whether to invest in paid traffic, SEO content, or catalogue expansion, based on what your real margin and operational capacity can actually support, not on how healthy the store looks in a screenshot.