About
Subscribe

From experimentation to execution: Can AI deliver real business outcomes?

Johannesburg, 09 Jul 2026
Carel du Toit, CEO, Mint Group.
Carel du Toit, CEO, Mint Group.

There is no shortage of noise around AI. Mordor Intelligence predicts the growth of the generative AI market to reach $28.45 billion in 2026, rising to $126.6 billion over the next five years.[1] There has been a staggering influx of tools, solutions, systems and platforms – research from Startus shows that there are nearly 22 000 companies active in generative AI and more than 16 400 funding rounds associated with companies focused on innovation intelligence.[2] While these numbers are constantly fluctuating, there are hundreds of new GenAI start-ups and solutions and predictions and promises and they are all putting companies under a lot of pressure to do something. To achieve with AI.

Most companies are already experimenting with the technology. Asking questions around how it really lands around transformation and whether or not it delivers on productivity or security or any of the metrics it has promised. This uncertainty is compounded by the adoption problem. How can companies adopt at scale while ensuring it is driving business?

The answer is twofold. On one hand, it is essential to identify which trends will deliver tangible business outcomes and value. On the other, it is key to align this adoption with governance as this is key to mitigating unnecessary risk and ensuring the business doesn’t face unexpected or unnecessary problems down the line.

In South Africa, companies are sitting in the middle of the AI conversation as many are still maturing their cloud investments and infrastructure. While the country certainly isn’t in its infancy, it is behind other regions where AI and cloud have become more entrenched in the business. The UAE, for example, is sitting in the 70th percentile in the Microsoft Global AI Diffusion Report and is in the lead globally, but South Africa isn’t there yet. However, the country is maturing and progressing at a rapid pace, and this means that the conversations around value are now essential. Companies must ask where the value is coming from, what they are solving for and how AI is going to drive value before they invest.

The step-change comes in as companies move from experimentation to execution. It’s not about making the business process better, it’s about redesigning the business process and rethinking it completely. The value of AI lies in changing the shape of the process in a way that measurably reduces cognitive or administrative load. It’s not putting AI into an existing workflow and tweaking one or two elements, effectively decorating business processes with AI capabilities.

Another consideration is, of course, data sovereignty. As AI becomes increasingly ubiquitous, so do conversations around where the data fed into the AI is stored. Microsoft has prioritised data sovereignty in its offering because the company has recognised how political landscapes have different expectations and requirements. With data centres in South Africa, the company solves half the sovereignty problem with data residency but not the full story of compute across where the apps and verification reside. As companies ask this question more and more, Microsoft is working with them to create 100% sovereign capability. And this is across both the public and private sectors. For South Africa, there is also the skills conundrum and a growing need for local talent and models that have the ability to understand local languages. The value created by this investment into local solutions means that innovation stays within the country rather than being exported with the talent.

As these factors continue to evolve and adapt to the needs of the business, it is going to become increasingly easier for companies to execute on their AI strategies in meaningful ways. The key is to remember that AI doesn’t transform the business; redesigned processes and properly governed systems do. That is the real work behind the value felt by AI implementation and this is where the Mint Group helps you to begin.

[1] https://www.mordorintelligence.com/industry-reports/generative-ai-market

[2] https://www.startus-insights.com/innovators-guide/generative-ai-report-key-insights/

Share