HEATH MUCHENA: Hunting unicorns with AI: the new frontier of start-up investing

Imagine this: you’re an investor with a sharp suit, a sharper eye, and an appetite for the next big thing. But instead of relying on gut instinct you’re armed with something infinitely more powerful — a tool that can analyse millions of data points, read market sentiment like a seasoned economist, and spot the next billion-dollar start-up — the next unicorn — long before anyone else can.

This isn’t the future; it’s happening right now. Artificial intelligence (AI) is transforming how investors identify and back high-growth start-ups, revolutionising the very nature of capital allocation. In Africa’s dynamic start-up ecosystem, where the stakes are high and the potential limitless, AI could be the golden ticket. 

Unicorns are rare, magical creatures — start-ups valued at more than $1bn. But unlike the mythical beast these are real and highly coveted by investors worldwide. The question is, how do you spot them early? AI models are doing just that by processing huge amounts of data in real-time, capturing everything from funding patterns and market trends to competitive dynamics and social sentiment.

Take for instance the dramatic shift in the African start-up funding landscape in 2023. Kenya, often overshadowed by heavyweight neighbour Nigeria, leapt ahead to claim the top spot, securing $800m in funding — about 28% of Africa’s total start-up investments that year​. 

AI tools, analysing key indicators such as market maturity, founder networks and regional funding shifts, could have predicted this surprising outcome, positioning investors to capitalise on Kenya’s surge before it became obvious. 

Think of AI as a digital detective piecing together a start-up’s DNA. What makes a unicorn a unicorn? It’s not just about the obvious markers such as revenue growth or market share. It’s also about the more subtle clues — team dynamics, consumer engagement, social media sentiment and even geopolitical developments. 

AI models scour millions of data points from a variety of sources — financial news, investor reports, social media chatter and more — to create a comprehensive profile of a start-up. Take fintech, for example, the reigning king of Africa’s start-up ecosystem. In 2023 fintech start-ups claimed a whopping 45% of the total funding on the continent​. 

AI could analyse this data, alongside non-traditional indicators such as public sentiment, transaction volumes and even mobile phone usage data, to predict which fintech start-ups are on track to become unicorns. 

Picture this: you’re an investor and your AI tool flags a small, Nairobi-based cleantech start-up that isn’t on anyone’s radar yet. Your tool identifies that this start-up — let’s call it EcoWave — is attracting attention from influential social media accounts, has a founder with a track record of successful exits and has just formed a strategic partnership with a global renewable energy giant. A week later EcoWave closes a $50m funding round, and you’re already on the inside. That’s the power of AI-driven insights. 

AI doesn’t just make capital allocation smarter; it makes it more inclusive. Start-ups in less visible markets or unconventional sectors often struggle to attract early investment. But AI is changing this. It breaks down geographical and sectoral biases by spotlighting high-potential start-ups regardless of where they’re located or what they do. 

For example, while Nigeria saw a dip in overall funding in 2023 — down to $410m from more than $1bn annually in previous years — other sectors outside fintech started gaining traction. The energy sector, particularly cleantech, emerged as a dark horse, attracting investments across Africa​. 

AI tools picked up on this early, flagging opportunities traditional investors might have overlooked, such as SunKing’s $130m securitisation deal in Kenya or Nuru’s $40m Series B round in the Democratic Republic of Congo. 

AI is powerful, but it’s not a crystal ball — yet. Think of it more like a partner in a dance, where human intuition and machine intelligence move in tandem. Successful investors don’t just rely on data; they blend it with gut feel, market experience and a dash of bold risk-taking. 

But AI does provide a critical edge. It can process and synthesise enormous volumes of data faster and more accurately than any human could, giving investors the ability to pivot quickly and decisively. In volatile markets such as those in Africa, where economic, political and social dynamics can change on a dime, this agility is priceless. 

While AI is a powerful ally, it’s not without its challenges. Data quality and availability remain hurdles. In many African markets data is still fragmented or incomplete, which can skew AI predictions. There’s also the issue of bias — AI models are only as good as the data they’re trained on. If that data reflects historical inequalities or biases, the AI could end up perpetuating them. 

Then there’s the trust factor. As much as we love the idea of machines making perfect predictions, there’s still a learning curve in trusting AI models. This is especially true in a domain such as venture capital, where gut instinct and personal networks have traditionally reigned supreme. 

AI is more than just a tool — it’s a paradigm shift. It’s changing how we allocate capital, identify opportunities and even think about risk. For Africa’s burgeoning start-up ecosystem this could mean more than just a wave of new unicorns. It could democratise access to capital, fuel innovation across a wider range of sectors, and help close the funding gap that so many promising start-ups face. 

So, next time you hear about a start-up achieving unicorn status remember: it might not be magic. It could well be an AI model quietly working behind the scenes, putting the pieces together long before the rest of us catch on. As Africa continues to rise as a start-up hub, these digital tools will be crucial in guiding where and how the next big bets are made. 

• Muchena is founder of Proudly Associated and author of ‘Artificial Intelligence Applied’ and ‘Tokenized Trillions’.