Business School - Ideas Are Cheap. The Founders Who Win Are Already in the Market

The most expensive habit in African startup life is still the polished plan that never meets a customer. Founders workshop the deck, refine the TAM slide, and wait for the product to feel finished. Meanwhile a competitor ships a clumsy first version, collects the insults, fixes the one thing that actually blocks payment, and is three learning cycles ahead before the first team leaves the starting blocks. Ideas are cheap because they have not yet collided with reality. Execution is the collision. That is the winning angle: not recklessness, but a bias for tests that produce data. The first version is supposed to be imperfect. If it is perfect, you waited too long.

Why action beats thinking

Thinking feels like progress. It is private, reversible and flattering. Action is public, expensive and embarrassing. That is why action teaches. A business plan cannot tell you whether a Durban salon will trust a Bluetooth card reader, whether a Dakar fishmonger will walk to a blue kiosk, or whether a Lagos merchant will paste seven lines of code into a checkout. Only a live transaction can. The founders who look “lucky” usually ran more cheap experiments than their peers. They treated launch as research, not as a graduation ceremony.

Africa punishes delay more than most markets. Payments fail. Agents disappear. Regulators move. Customers switch on price. A twelve-month planning cycle is a year in which the problem you designed for has already changed. The impressive African companies of the last decade were not the ones with the most elegant theses. They were the ones that put an ugly product in front of users and let the users rewrite the thesis.

Paystack: start with a party trick, then stay in the loop

Shola Akinlade and Ezra Olubi did not begin with a 40-page strategy. Akinlade, then a software engineer, found he could charge a debit card from his laptop and started doing it as a demonstration. Friends brought cards. The pain was obvious: Nigerian businesses could not accept money online the way a Stripe merchant could. In 2015 they built underground with a waitlist of about 300 merchants, iterating features against real complaints rather than imagined ones. Y Combinator rejected them once. They applied again, became the first Nigerian company into the programme, and launched publicly in January 2016 with a blunt promise: go live in about 30 minutes.

That promise was execution, not branding. The product had to be simple enough that a small shop could integrate it without a bank committee. Volume followed the integrations. By mid-2017 Paystack was processing around a billion naira a month. Stripe led the Series A, then bought the company in 2020 for more than $200 million. The lesson is not “get acquired.” It is that the year of quiet merchant loops beat another year of planning the perfect gateway. Ten years on, Paystack is still shipping — a microfinance bank, a consumer app, a holdco — because the habit of putting product in front of users never stopped.

Yoco: 500 merchants before the brand campaign

In South Africa, Katlego Maphai and co-founders saw a measurable gap: most consumers carried cards, almost no small merchants could accept them. They did not wait for a finished hardware story. A prototype reader was with pilot merchants by late 2014. About 500 businesses ran the beta. Only then did Yoco launch properly in 2015 with a phone-attached card machine.

The first lesson was distribution, not design theory. Shop floors taught them which cables snapped, which onboarding steps died, which settlement delays killed trust. The second lesson arrived in 2020. Yoco was an offline POS business when lockdowns shut the floor. Because the team had already been building an online gateway, they could push card-not-present tools fast enough that merchants kept selling. Iteration is not only the first launch. It is the willingness to use the next shock as another test. Yoco’s later platform — invoices, QR, e-commerce plugins — is what you get when hardware teams keep listening after the first reader works.

Wave: a simpler product, launched where the cash already moved

Wave did not invent mobile money. Orange Money already owned the rails in Senegal. Drew Durbin and Lincoln Quirk, coming off Sendwave, launched Wave in Senegal in early 2018 with a narrower bet: an app, QR codes instead of clunky USSD, deposits and withdrawals that did not punish the user, and a 1 percent send fee against incumbents charging several times that. They put blue kiosks where cash already lived and let price plus ease do the arguing.

Within a few years Wave had millions of Senegalese users and later took a meaningful share of UEMOA electronic-money volume. Valuation followed usage, not the other way around. The execution lesson is brutal and useful. They did not out-plan the telco. They out-shipped a cheaper, clearer behaviour and staffed the agent network as if it were the product — because it was. A founder still polishing a “pan-African wallet thesis” in 2018 missed the year Wave spent learning which kiosk locations converted.

Moniepoint: when the customer is too slow, sell the product yourself

TeamApt, the company behind Moniepoint, first built white-label agent-banking software for Nigerian banks. The banks moved too slowly to put that software into the informal economy at scale. The founders did not write a longer memo. They took the same stack and deployed it themselves to agents and small businesses. Uber’s Nigerian driver payouts became an early public proof that the pipes could take volume.

That is a pure action pivot. The original idea — sell to banks — was not wrong. The timing of the buyer was. Waiting for the institution would have killed the company. Serving the end user produced the business that later became one of Nigeria’s defining financial platforms. Lesson: if the channel you planned will not move, the product may still be right. Change the channel in the market, not on the whiteboard.

What these stories share

None of these teams launched finished. Paystack’s first checkout was a workaround. Yoco’s first reader was a prototype. Wave’s first network was a beachhead, not a continent. Moniepoint’s first customers were not the ones in the original slide. Each used a constrained first version to buy information: who pays, what breaks, which fee people will walk across town to avoid. The “overtaking” the armchair founder fears is rarely a better idea. It is more cycles.

Failure is part of the same method. Plenty of African startups launched fast and still died — thin unit economics, fraud, regulation, a market that looked large in a deck and tiny in cash. Action is not a guarantee. It is the only way to discover the fatal assumption while you still have runway. The founders who look disciplined are often the ones who scheduled the embarrassment early.

Practical steps: leave the starting blocks

Define a test that can fail this month. Not a brand. A transaction: one merchant live, one agent float turning, one user sending money twice. If the test needs a committee, it is still a plan.

Cap the build. Give the first version a date and a “good enough” list of three features. Everything else is a sequel. Perfect onboarding copy is not a feature if nobody has paid.

Put a number on feedback. Paystack’s 300-merchant waitlist was a queue of conversations. Yoco’s 500 beta devices were 500 arguments with reality. Write down what broke. Ship the fix. Repeat weekly, not after the next offsite.

Sell before you scale the story. Wave’s fee was the pitch. Yoco’s gap — cards in pockets, no machines on counters — was the pitch. If you cannot say the offer in one sentence to a busy shopkeeper, you are still thinking.

Change the channel when the buyer stalls. Moniepoint’s bank pivot is the template. A slow customer is data. Do not interpret it as a reason to polish the same deck.

Keep a kill rule. If two consecutive tests produce no repeat use, stop adding features. Change the problem or stop. Momentum is not activity. It is evidence that someone came back.

Protect the doing hours. Strategy sessions expand to fill the diary. Put customer calls and live deploys on the calendar first. The plan can happen after the receipts.

Ideas will keep arriving. They always do. The scarce resource is a founder willing to let the first version be wrong in public, then use the bruise. African markets will not wait for the perfect plan. They will, sometimes, wait ten minutes for a product that takes the money.