Every week brings another tool that promises to change everything. For most small and mid-sized businesses in Africa, the hard part is not finding AI. It is working out where it actually belongs.
Start with the leak, not the tool
The businesses that get value from AI rarely begin with AI. They begin with a clear-eyed look at where money and time leak out of the business. A day lost every week pulling the same report. Invoices re-keyed by hand. Customers who go cold because nobody followed up. Find the leak first, and the question stops being what can this tool do and becomes what is this costing me now.
Once you can put a number on the pain, you can judge any tool honestly. If a process wastes ten hours a week and a machine can do it in minutes, the case makes itself. If it does not waste much, no amount of clever technology will make it worth the effort.
Three places AI tends to earn its keep
- Repetitive back-office work, where the same steps repeat hundreds of times a week and the rules are clear.
- First-line customer questions, where most of the volume is the same handful of queries asked in different words.
- Reading and sorting documents, where people spend hours pulling numbers out of PDFs, forms and messages.
None of these are glamorous. All of them are quietly expensive, and all of them are within reach of tools that already exist today.
And where it usually does not
AI struggles where judgement, relationships and trust do the real work. It will not replace the conversation that closes a deal, the instinct of an experienced operator, or the accountability a customer wants from a person. Point it at those and you get an expensive disappointment.
If AI is not the answer, the honest thing is to say so. We would rather lose a sale than sell you something that will not move your business.
Test it for the price of a week, not a quarter
You do not need a big budget to find out. Pick the single highest-value leak, run a small pilot against your real data, and measure it against the number you started with. If it works, scale it. If it does not, you have spent a little to learn a lot, and you still own the insight.
That is the whole game. Start with the business, prove it cheaply, and let the results decide what comes next.