This is an exercise I am starting in order to get an understanding of existing problems (or opportunities) in the African AI landscape.
I intend to conduct project discovery research in order to draw inspiration from a hybrid of problems that plague me in my everyday life, as well as (internal/external) communities and stakeholders of interest to me.
My Project Discovery exercise
- 10 problems that I encounter frequently
- 10 problems that Africans encounter (Conduct a 0/1 knapsack survey to identify pain points and top ranking possible services)
- 10 problems that businesses encounter (Conduct a 0/1 knapsack survey to identify pain points and possible services)
- 10 problems in the domain of Computer Science (databases, data mining, and/or artificial intelligence)
- 10 problems big tech companies are solving
- 10 problems unicorn companies are solving
- Market research and analysis
One of the key goals of the startup discovery phase is to make sure there’s rising user demand for a solution. The outcome of the research should also enable you to answer these 3 questions:
- Should you really do this?
- Can you do this better?
- How much resources do you need?
Resources: Articles I read while coming up with this research plan


Notes/Insights
- 42% of startups fail as there’s no real need for their product. 29% fail as they incorrectly estimate the budget. 19% fail because they get outcompeted or for other reasons.
- Vertical > Horizontal model: Unlike a horizontal approach, which attempts to cover as many fields as possible, a vertical SaaS strategy focuses on specific needs within an industry and allows getting closer to the potential client.
Hottest SaaS trends:
- AI
- Blockchain
- Vertical SaaS - Offer complex solutions for a specific niche instead of targeting a few industries
- Pay per use - PPU model looks more attractive to clients and can stimulate to use your product. Conduct a financial analysis before settling on a pricing strategy.
- Invest in satisfying your customers more than constantly looking for new ones
In the next post, I will share my findings. Stay tuned!