Converting AI Hype into Tangible Business Outcomes: A CIO's Guide

AI priorities in the enterprise aren’t always fully understood. Technologies and investment can work together or conspire against each other in equal measure. The approach for IT leaders in South Africa, for instance, is to proceed with caution.

Credit: PeopleImages.com - Yuri A / Shutterstock

In Foundry’s AI Priorities Study 2025, businesses reveal they’re allocating more money to AI projects than ever before, with nearly half of organizations now dedicating budget for AI projects, up from 36% in 2023. Plus, they’re allocating almost a quarter of IT spend to AI initiatives.

While some are taking a more measured approach than others, the consensus among South African tech leaders in particular is that AI investments must deliver tangible results, and that AI budgets have to be channelled strategically across the business. At financial services group Sanlam, it’s less about upping AI budgets, says group CIO Theo Mabaso, and more about allocating more of the budget toward high-impact areas. “After a period of experimentation, launching multiple AI outcomes and learning by doing,” he says, “we’ve figured out where the objective AI value is concentrated, and have thus narrowed our focus to delivering fewer but higher impact AI outcomes that deliver ROI for the group.”

For example, it’s understood that gen AI assists productivity and enables employees to focus on high-value tasks. And Sanlam has seen time saving and benefits through the likes of GitHub CoPilot, and using LLMs to review lengthy contracts against standard terms and conditions. “However, we recognize that productivity doesn’t always drive objective metrics, so we’re critical of where we deploy such technologies,” he says. “We look for an area where a variable can be affected to create an objective downstream effect, as opposed to large-scale deployments of productivity tools that deliver soft benefits tied to multi-year contracts.”

Werner Leithgöb, Lactalis Southern Africa

Lactalis

Werner Leithgöb, IT director of dairy manufacturer Lactalis Southern Africa, agrees. He believes the elevated hype around AI only scratches the surface of what these tools can do. “When the Amazon Alexa first came out, everybody thought it was so awesome because you could interact with the devices in your home ecosystem using voice commands. But I guarantee that 99% of consumers who have one use it purely to stream music. So, essentially it’s a glorified speaker. For me, AI is very similar. In order to utilize this technology properly, you have to be very deliberate and define clear use cases for it so people can see what value it adds.” As part of being deliberate, he says, you must regularly use the technology so it becomes part of how you work.

Maximizing potential

As use cases evolve, just experimenting with AI for the sake of innovation doesn’t cut it anymore, says Jenny Mohanlall, senior director for IT at DHL South Africa. “It’s about leveraging it to solve real problems, drive efficiency, and create meaningful impact,” she says. For most organizations, AI priorities are focused on enhancing the customer experience and revolutionizing operations and ways of working.

Jenny Mohanlall, DHL South Africa

DHL

The findings of Foundry’s AI Priorities Study support this, highlighting that IT departments are more readily partnering with other business units, like customer service and marketing, to ensure AI adoption aligns with broader business goals. By leveraging AI tools to streamline and improve something like customer support, it’s possible to free up human agents to tackle more complex issues. And by using sentiment analysis tools to dig into customer feedback, businesses can identify areas for improvement and handle issues before they escalate. “Some even take it a step further with predictive support, where AI anticipates customer needs and proactively addresses issues before they escalate,” says Mohanlall. “Imagine a scenario where a chatbot not only answers questions but also predicts what a customer might need next based on their behavior. This is the kind of innovation that’s transforming customer service.”

AI in action

As part of their efforts to democratize financial guidance, build financial literacy, and support financial decisions in a simplified way, Sanlam developed an AI product called Coach. The first iteration was a free tool available 24/7 designed to enhance financial literacy and empower users to manage their credit more effectively. 

Theo Mabaso, Sanlamup

Sanlamup

The technology behind Coach included advanced data and analytics, and utilized 40 LLMs to provide hyper-personalized financial guidance to users so they can make informed financial decisions based on their unique context, says Mabaso. “In its first few months, the product facilitated 45,000 unique conversations and thousands of loan applications at a conversion rate three times higher than that of web or WhatsApp channels,” he says. The second iteration of Coach, released a few months later, delivered a gen AI-driven money personality assessment designed to help clients uncover their unique financial behaviors, enabling Sanlam to tailor financial advice. It also empowers their advisors to offer customers the best solutions to help them meet their financial goals. 

Speaking broadly about how AI is transforming logistics, Mohanlall says AI initiatives are redefining what’s possible by helping the industry to better manage unpredictability. “For example, by analyzing historical data, weather patterns, traffic conditions, and even global events, businesses can predict demand and optimize delivery routes in real-time.”

For Leithgöb, the journey is just starting but he isn’t worried about falling behind because he understands the importance of doing the necessary groundwork for AI success. “We’ve got a bunch of Copilot licenses and we’re testing it with certain users, encouraging them to play with it, try out different prompts, and educate themselves on how to use it,” he says. “Let’s get people familiar with it while we lay the necessary data foundations.” Right now, he admits, getting their data in order is the main priority. “AI thrives on data, so if your data is a mess, your AI is also going to be a mess,” he adds. “If you want to generate useful insights from vast amounts of data, you need the right tools but you also need the right data.”

For these tech leaders, the goal is to use AI to minimize manual and repetitive tasks, streamline processes, and unlock fresh insights. But ultimately, all of these efforts should come together to help the business cut costs, increase sales, and lift profits.