E-Commerce
What an ecommerce website costs in South Africa, and where the money goes
The storefront is the part of an ecommerce website you can see, and it is the smallest part of the bill. On the Shopify builds we have scoped for medium and large South African businesses, design took about a quarter of the budget. Most of the rest went on work a customer never sees: moving data, connecting the store to the systems that run the business, and proving the whole thing works before it takes an order.
Simon Bestbier
5 min read
That is why two quotes for what looks like the same store can be a different size of project. If you are budgeting for a build, or trying to explain a quote to your finance team, the useful question is what sits underneath the storefront and how much of it your business needs.
Most of the businesses we see are already selling online
Most of the businesses that come to us are on WooCommerce or Magento. The store works, but it is slow, fragile or expensive to keep running, and every change takes longer than it should. A smaller group sell offline or B2B and are building their first serious online channel. A few are on Shopify already and have outgrown a theme-based store.
The starting point changes the shape of the budget. A replatform carries years of products, customers and orders that have to come across. A first serious channel usually needs the store to talk to an ERP from day one. A business outgrowing its theme needs custom design and the integrations it put off the first time.
Where the budget goes on a scoped build
These are approximate shares from our recent Shopify builds. Every project moves them, but the pattern holds.
Design: about a quarter of a build with a custom design.
Integration: about 40% of a build that connects to an ERP, stock system or other back-office platform.
Data migration and go-live: about 10 to 15%.
Quality assurance, delivery management and contingency: about 25 to 30%.
The last line is the one buyers question most, because it reads like overhead. It is the work that stops a store going live with stock it does not have, prices that are out of date or orders that never reach the warehouse.
A cheaper quote usually leaves out the testing. The cost of that shows up after launch.
Testing an ecommerce store means testing every path an order can take, with the client's own team signing off before launch. On a build with integrations, that work grows with every system the store connects to.
Data migration is the line buyers underestimate most
Clients tell us the data is ready. It rarely is. Product data that has built up over years on WooCommerce or Magento carries duplicate products, variants set up three different ways and products with missing images or attributes. Shopify is stricter about how products and variants are structured, so someone has to fix the mess before it comes across.
Customer accounts are the second surprise. Passwords cannot move between platforms, so every customer has to reset their password or be invited back to the new store. That needs a plan for the emails, the timing and the customers who do not respond, and it is part of the migration budget.
Order sync with the ERP is where integration budgets go
Clients put custom integration second on the list of things they underestimate, and order sync with the ERP third. In practice they are the same problem. Four things make it harder than it looks.
The ERP was never built for real-time ecommerce. Many run batch jobs overnight, have no API or have one that cannot take the load of a busy store.
Every edge case has to round-trip. A partial shipment, a return, a cancellation, a back-order or a split payment has to update both systems correctly, every time.
An integration is priced on how well the ERP behaves, and most businesses only find that out once the work starts.
Stock has to be right. When the ERP and the store disagree, the store oversells, and the problem multiplies across warehouses and sales channels.
Nobody owns the ERP data. The rules for pricing, stock and fulfilment often live in one person's head or in years of workarounds, so they have to be discovered during the build.
How to keep the number honest
There are four things we do on every build of this size.
A paid discovery phase first. Before we quote the build, we look at the data, the ERP and the integration rules. That moves the biggest risks from the build into a short, fixed piece of work, and we base the build quote on what we find.
Phasing. The core store launches first. Integrations and B2B features follow in later releases, once the store is trading and the business knows what it needs.
Data clean-up before migration. The client's team or Realm fixes the product and customer data as a separate piece of work, so the migration moves clean data once.
Fixed scope with a change process. We price additions when they come up, so the budget you approved stays the budget you are working to.
The quote may look higher on paper. The number you approve is the number you pay, which is what a finance team needs from an ecommerce budget. For the running costs after launch, our Shopify pricing calculator shows what the platform and transaction fees come to each month in rand.
Start with the parts you cannot see
If you are planning a replatform or a first serious online channel, the most useful thing to know before you ask for quotes is the state of your product data and how your ERP handles orders and stock. Those two answers set most of the budget. If you would like help finding them, we run scoping calls from our Cape Town office for businesses across South Africa, and you can read more about how we approach ecommerce development.
