Start with what you can already sell
The usual advice is to look for a gap in the market. It is less useful than it sounds, because a gap you cannot fill is not an opportunity. A more reliable starting point is yourself. Ten years in a function gives you knowledge of how an industry works and where it fails, a set of contacts, and credibility with a specific kind of buyer. A new business has none of these on day one, and you already own them.
One exercise helps more than most. Write down the problems you have solved repeatedly in your working life, for your employer or for the people around you. Problems that are easy for you are often ones others cannot solve at all. That gap is where a business usually starts. The article How to Find a Business Idea in South Africa works through this in more detail.
Four things decide fit
If you are still working out whether to leave employment at all, that is a separate decision with its own checks. Read should I leave corporate and start my own business first.
Eight types of business, and what each demands
Almost every business a corporate professional starts falls into one of eight types. The type matters more than the specific idea, because it sets how much capital you need, how long the first income takes and which mistakes are likely.
| Type of business | Capital needed | Time to reliable income | Where new owners get caught |
|---|---|---|---|
| Consulting and advisory | Low | Weeks to months if you have a network | Positioning so broad that no one can tell what you do, and a client ramp that takes longer than your savings last. |
| Agency and professional services | Low to moderate | Months | Underpricing, and hiring early: staff add fixed cost and labour obligations before revenue is steady. |
| Digital products and online | Low to moderate | Slowest of the eight: often 12 to 18 months of steady output before results | Overestimating the reachable South African audience for a niche, and pricing the product too low. |
| Physical products and e-commerce | Moderate | Months, with cash tied up in stock | Paying for stock before customers pay you, and minimum order quantities that commit capital before demand is proven. |
| Manufacturing | Very high, lower if you start with a co-packer | The longest cash cycle: months from raw materials to payment | The heaviest regulation of the eight, and demand that is harder to prove than production. |
| Retail and food service (not franchised) | High | Months, after fit-out | The lease: typically three to five years with no break clause and often a personal surety. Location decides most of the outcome. |
| Franchise | High | Months, after fit-out | Royalties come off turnover, not profit, so the unit needs enough margin left after royalty, advertising levy, rent and staff. |
| Property and accommodation | High to own, low to manage other people’s property | Rental income can start quickly, but a single property rarely replaces a salary | Confusing capital growth with cash flow, and local rules on short-term rental. |
How to match a type to you
Read the table against your own answers to capability, capital, time and risk. These pairings are a starting point for a shortlist, not a verdict.
- Under six months of runway: consulting, advisory or professional services, where you can invoice within weeks.
- Little capital and an idea that needs stock: start with a service that earns while you test the product, or take orders and payment before buying inventory.
- A wish to build something you can eventually sell: digital products and agencies can run on systems and people. A consulting practice depends on you.
- Plenty of capital and a low appetite for building from nothing: an established business or a franchise, where the model already exists.
- Strong experience in one industry: a business that serves that industry, after you have checked your employment contract.
A type that suits you on paper still has to pass a test against real customers. The next section covers how.
Choices that look good and fail the fit test
- Choosing from a trend instead of from what you can sell. A fashionable category still needs a customer who will pay you.
- Picking a stock-heavy business on a service-sized budget. Minimum order quantities can commit most of your capital before you know whether anyone wants the product.
- Assuming an interest is a market. Enjoying something does not mean enough people will pay for it at a price that works.
- Choosing a type that takes longer to pay than your runway lasts. A content business that needs 12 to 18 months of output cannot support a household on six months of savings.
Narrow to two to four, then test
- Write each candidate as one sentence: who the customer is, what problem you solve for them and what they pay.
- Score each candidate against capability, capital, time and risk. Drop any that clearly fail one.
- Work out the break-even volume for each survivor in the break-even calculator. If the daily sales it demands look unrealistic, drop it.
- Test the strongest one or two with real customers before you build anything. The guide How to Test a Business Idea in South Africa covers how to tell demand from polite enthusiasm.
- Choose the one that survives the test, not the one you liked best at the start.
The Launchworks evaluation method follows the same sequence in two stages: discovery to find the direction, then a stress-test against customer, competition, money and personal fit.
Where to start
Choose the description closest to where you are.
I want my own business but I do not know what
A guided conversation that works from your skills, your capital and the life you want, and narrows the field to two to four directions. Each comes with an honest read on capital, time to revenue and risk.
Find your direction →Have a shortlistI have an idea and need to know if it holds up
The stress-test works through customer, competition, money and personal fit. It ends with the one number that decides whether the idea is worth pursuing.
Test your idea →The methodI want to see how the evaluation works first
The two-stage method in full: discover the direction, then stress-test it before you commit money.
Read the method →Considering buyingI would rather buy something that already trades
Buying removes the idea risk and adds price and debt. The six-stage guide is written from the buyer’s side, not the broker’s.
Read the guide →If you are weighing starting against buying or franchising, read start, buy or franchise. If a franchise is on your list, the buying a franchise guide covers the disclosure document, the numbers and the agreement. Once you have chosen a direction, the knowledge base has 49 South African guides to registration, tax and compliance.
Further reading
- What Small Business Can I Start in South Africa?
- How to Find a Business Idea in South Africa
- The Business Idea That Passed Every Test and Still Didn’t Get Built
- How to Test a Business Idea in South Africa
- Why Most R5 000 Business Ideas Fail Before They Reach a Single Customer
- Launchworks Opportunity Report 2026
Frequently asked questions
What business should I start if I have never run one?
Start from what you have already done professionally, not from a trend. A service that uses your existing expertise, such as consulting, advisory or professional services, needs little capital and can earn within weeks, which lets you learn to run a business while it pays you. Product, retail and manufacturing businesses add stock, premises and a longer cash cycle to the learning.
What is the best business to start in South Africa?
There is no best business, only the best fit for your skills, capital, time and risk tolerance. Service businesses usually pay soonest and need the least money. Manufacturing, retail and property need the most capital. The right answer for you comes from testing a shortlist, not from a ranked list of ideas.
What business can I start with little money in South Africa?
Service businesses are the usual answer: consulting, advisory, agency work, training and coaching need little more than your time, a laptop and a contract. A product business can start small if you take orders and payment before you buy stock. Whatever you choose, work out your break-even volume before you spend anything.
Should I start a business in the industry I work in?
Often yes, because your credibility, contacts and knowledge of where the industry fails are the assets a new business lacks. The catch is contractual. Check your employment contract for restraint of trade and conflict-of-interest clauses, and do not compete with your employer or use its clients or information while you are still employed.
How do I know if a business idea is good?
A good idea survives four checks: a specific customer who will pay, a competitive set you can realistically beat, numbers that leave room for you to be paid, and a fit with your own capital and time. Run the idea through all four before you spend money. The stress-test works through them with your own figures.
How many business ideas should I test at once?
Shortlist two to four, then test the strongest one or two properly. Testing more than that spreads your time thin while you are employed. Testing only one makes it hard to walk away from a weak result, because you have nothing to compare it with.
Is it better to buy a business than start one?
It depends on whether your constraint is ideas or capital. Buying removes the idea risk and the slow start, and adds a purchase price and the work of verifying the seller’s numbers. Starting costs less up front and takes longer to reach income. A franchise sits between the two, with its own fees and obligations.
How much money do I need to start a business in South Africa?
It ranges from close to nothing for a service business to millions of rand for manufacturing, and the figure that matters is your own. Work backwards from the monthly income you need, add the months before the business pays you, then add what the business type ties up in stock, fit-out or deposits. The runway and break-even calculators are free.