The question I get more than any other, from owners across South Africa, is some version of the same thing: which AI should I buy. Which tool, which chatbot, which platform. I understand the question. Everyone can feel that something has shifted and nobody wants to be the last one still doing it the slow way.
It is still the wrong question. And I say that as someone who sells the answer.
Buying the tool first automates the mess
Here is what I have watched happen, more than once. An owner buys an impressive piece of AI. They wire it into their business. And a month later nothing is better, sometimes it is worse, because the tool faithfully sped up a process that was broken to begin with. AI does not fix a bad workflow. It runs it faster and more confidently, which is not the same as fixing it. If your intake is chaos, an AI that handles intake gives you faster chaos.
The tool is almost never the constraint. The constraint is that most businesses have never actually mapped how work moves through them. The knowledge lives in three people's heads and a WhatsApp group. Nobody has drawn the picture. So when you drop AI onto that, you are automating a process nobody fully understands, and you find out where the holes were only after they are moving at speed.
What an ops autopsy actually is
An operations autopsy is exactly what it sounds like. Before you buy anything, you lay the business out on the table and look at how it really runs, not how the org chart says it runs. Where does a lead actually enter. Who touches it. Where does it wait. Which steps are a person copying something from one place into another. Where does the same information get typed three times. Where do things quietly go to die.
When you do that honestly, two things happen. You find the real bottleneck, which is usually not where anyone guessed. And you find the handful of tasks that are pure mechanical repetition, the ones that are genuinely worth handing to a machine. That is the shortlist worth automating. Everything else is noise you would otherwise have paid to speed up. I wrote a full walk-through of what this looks like across four hours in what a 4-hour Business Autopsy actually looks like.
I learned this the expensive way
I am not preaching from theory here. When I started NexBDM, I built the machine before I had proof anyone wanted what it made. Registration, website, logo, the whole setup, all of it ready before I had spoken to a single customer. I built the answer before I understood the question, and I have written honestly about the morning I finally turned the tools on my own business and did not like what I saw.
The AI version of that mistake is everywhere right now. Buying the capability before understanding the job. It feels like progress because money left your account and something impressive arrived. But spending is not the same as solving. The autopsy is the cheap step that stops you buying the expensive wrong thing.
The three things it surfaces every time
When I sit inside a business and map it, the same three things come up almost without fail.
First, a bottleneck nobody had named. The owner is convinced the problem is sales, and the real jam is a single approval step that adds two days to everything.
Second, duplicated work. The same fact entered by hand into three different systems, because the systems were bought one at a time and never made to talk.
Third, a task that is a perfect candidate for automation and that nobody flagged, because it was so routine everyone had stopped seeing it.
None of those need AI to be discovered. They need someone to look. The AI conversation gets useful only after that look, because now you are pointing a powerful tool at a problem you have actually defined.
What I would tell someone facing the same thing
Do not buy the tool first. Map the work first. Spend the cheap hours understanding how your business really moves before you spend the real money making it move faster, because speed applied to the wrong process is just a more expensive version of the same problem.
The question is not which AI to buy. The question is where your business actually loses time, money and follow-through, and you cannot answer that from the brochure. You answer it by looking. If you want to see how we do that look, it is a Business Autopsy, and it is the first thing I would do before spending a rand on any tool, including mine.
Frequently Asked Questions
What is an operations autopsy? It is a structured look at how work actually moves through your business before you change anything: where leads enter, where they wait, which steps are manual, and where the real bottleneck sits. The goal is to find the work worth automating, not to sell a tool.
Why not just buy AI and get started? Because AI speeds up whatever process you point it at, including a broken one. If the underlying workflow is unclear or duplicated, automating it makes the problem faster and more confident, not smaller. Understanding the process first is what makes the tool pay off.
How is this different from a general business consultation? An ops autopsy is narrow and practical. It maps the actual flow of work to find bottlenecks and repetitive tasks, then produces a shortlist of what is genuinely worth automating. It ends in specifics, not a strategy document.
Do small businesses really need this? Smaller businesses need it most, because the process usually lives in a few people's heads and has never been written down. That is exactly the situation where dropping in AI causes the most surprises later.
More of the build in public at heinoux.nexbdm.co.za.