What an MVP costs to build in South Africa
What an MVP costs in South Africa depends on scope, who builds it and what you leave out. Here's how each one moves the price, with rough numbers.
"How much will it cost to build my app?" is the first question almost every founder asks, and the honest answer is "it depends". That isn't much help, so this post is about what it depends on, and how to keep the number down without building something you'll have to throw away.
Before the price, though, a bigger question: should you build it at all? Your idea deserves a fair test, whether that ends with building it or letting it go before you've spent too much money or time. Plenty of founders have put months and savings into an idea that a few hard questions would have exposed early.
The best way to get those questions is from someone who understands the market you're entering and is willing to play devil's advocate. That's hard to get from friends and family, who usually want to be encouraging. Look for someone who will be honest with you, and who knows the economics and the hidden practical problems of your field.
As a software engineer, I can test the technical side: whether it can be built, roughly what it takes, and where the risks are. I explain my free offer at the end of the post.
What an MVP actually is
A minimum viable product is the smallest version of your idea that a real customer would use, and ideally pay for. It is not a demo, and it is not version one of everything you've imagined. The goal is to learn, as cheaply as possible, whether people want it.
That distinction drives the cost more than anything else. Most MVPs that run over budget do so because they quietly became a full product.
The four things that set the price
Scope. How many kinds of user, how many screens, and how many things each one can do. An app where people sign up, do one thing and pay for it is a very different job from a marketplace with buyers, sellers and an admin team.
Integrations. Payments, email, a CRM, a supplier's API. Each is quick when it's well documented and slow when it isn't. Third-party APIs are where estimates most often go wrong.
Who builds it. An agency brings a team (project manager, designer, developers) and charges for it. A freelancer is usually cheaper and you talk to the person writing the code, but you depend on one person. Offshore teams can be cheaper again, with the time-zone and communication costs that come with them.
What happens after launch. Hosting, monitoring, fixes and the first round of changes once real users arrive. Budget for this from the start; it is never zero.
Rough numbers
A simple MVP
one kind of user
sign-up and log-in
one core feature that isn't too complex
payments through a payment provider
Roughly R50 000, and about a month.
A mid-sized MVP
two kinds of user (such as an administrator and an employee), or a marketplace (such as buyers and sellers)
one core feature that can be complex
an admin area
up to three integrations with external APIs
Roughly R100 000, and about two months.
Both are for building and launching; hosting and changes after launch are extra. Treat any figure you see online, including mine, as a starting point for a conversation, not a quote. Two apps that sound the same in one sentence can differ by months of work.
How to spend less without regretting it
Cut features, not quality. Leave out whole features rather than building everything badly. A small app that works is easier to grow than a big one that doesn't.
Use off-the-shelf parts. Payments, sign-in, email and admin screens don't need to be invented. On my own products I use a headless CMS (Directus) for the admin and data, which saves weeks.
Do some things by hand at first. If an admin task happens once a week, it can be a person following a checklist until the product proves itself.
Insist on a deploy pipeline from day one. It costs a little up front and saves a lot every time something changes, which in an MVP is constantly.
What I'd ask you first
Who is the first customer, what's the one thing they'll do, and how will you know it's working? If you can answer those, we can scope an MVP and put a real number on it.
I offer a free one-hour call to talk through your idea: I'll ask the hard questions about the idea itself, look at the technical side, suggest which features should come first, and flag anything that could cause trouble. Fill in my form to set up a time, or see how I work.