The local apps that survive tend to come from founders who had the problem themselves, because that is where the detail comes from. Booking a service, finding a trusted provider, managing a household, tracking a small business: the useful ones remove a specific, repeated annoyance rather than offering a platform for something general.
For anyone considering building one, the sequence below matters more than the technology choice.
Solve a problem you have had
Founders who lived the problem know the detail that makes a solution usable, including the parts a researcher would never think to ask about.
It also sustains you through the period where nothing works, which is longer than most people expect and is where most app projects end.
Validate before building
You can test demand without writing code. A landing page, a manual service delivered over messaging, or a spreadsheet operated by hand will tell you whether people want the outcome.
Building first and testing after is how founders spend their savings discovering that nobody wanted it. The manual version also teaches you the workflow you are about to automate.
Distribution is harder than development
Getting an app built is a solved problem. Getting people to install and keep using it is not, and it is where most of the money and effort actually goes.
Decide how people will find it before you build it. An app with no distribution plan is a project rather than a business.
Data cost and device reality
Applications that are heavy on data or require a recent device exclude a large part of the South African market. Light, tolerant of poor connections, and working on inexpensive handsets is a commercial decision, not a technical nicety.
The obligations that come with users
Any app holding personal information carries duties under the Protection of Personal Information Act, overseen by the Information Regulator: collect only what you need, say what you hold and why, secure it, and let people ask what you have about them.
Anything touching payments, credit or financial advice is separately regulated and needs authorisation before launch rather than after.
Frequently asked questions
What do lasting apps have in common?
A founder who had the problem themselves, which supplies the detail that makes the product usable.
How do I validate an idea without building it?
A landing page, a manually delivered version over messaging, or a spreadsheet run by hand. Test the outcome, not the software.
What is the hardest part of launching an app?
Distribution. Development is largely a solved problem; getting people to install and keep using it is not.
Why does app size matter in South Africa?
Heavy data use and recent-device requirements exclude a large share of the market, which is a commercial limit rather than a technical one.
What regulation applies to an app with users?
Data protection obligations for any personal information held, and separate authorisation for anything involving payments, credit or financial advice.
Further reading
Originally published in August 2018. Updated September 2026 to explain what app founders solve first rather than listing individual applications.
