I have never bought a bakkie without driving it on my own farm first.
I have bought software on a demo.
More than once.
(For readers outside South Africa: a bakkie is a pickup truck, except it is also a family car, a delivery van, a farm implement and a bar. We are not open to debate on this.)
You know the demo.
A clean system. Sample data. A company called something like Acme Industries.
Every scenario works perfectly. Every screen loads quickly. Every question has an impeccable answer.
But, you make biscuits in Cape Town.
They show you a water utility in Denmark.
And you are expected to make a solid buying decision on the strength of it.
Where the demo has gaps, there is a PowerPoint.
The PowerPoint never has gaps. That is what PowerPoint is for.
Nobody is lying. It is simply a bakkie on a showroom floor, with the engine off and the price tag turned the other way.
And at some point, usually around the forty-minute mark, somebody senior leans back and asks the only question that matters.
“Yes, but what would this find in our business?”
Silence.
Not the vendor. Not the consultant. Not the CIO. Nobody can answer that.
Because the demonstration was never about your systems. It was about Denmark’s.
So the organisation does what it has always done.
It commits first and finds out later.
A pilot. A phase one. A proof of concept, which is the polite name for paying to discover whether the thing works.
I have sat on both sides of that table.
I have given the Denmark demo. I have bought from one.
I have run the proof of concept that proved mostly that we should have asked a better question in month one.
For two decades I accepted the order of things.
Demonstration, then decision, then evidence.
That is how software is bought. Everybody does it that way.
Then, one day, it occurred to me that we had the order backwards.
The evidence already exists.
It is sitting in the system the executive is asking about.
The only reason the demo cannot answer the question is that nobody brought the ERP into the room.
Nobody buys a bakkie from a brochure.
You take it onto your own roads, your own corrugations, your own river crossing.
You listen to what rattles.
Then you decide.
Software should be no different.
The first thing on the table should be a finding from your own ERP estate.
Not a slide about what a platform could do in Denmark.
A measurement of what it did, on your system, before anyone signed anything.
Then the question changes.
Not “what would this find?”
But “what do we do about what it found?”
The first finding should arrive before the first commitment.
First published on LinkedIn, 10 November 2026.