AI may have short-term gains, but is a long-term play

Christopher Tredger
By Christopher Tredger, Portals editor
Johannesburg, 13 Mar 2024
Verushca Hunter, executive digital technologies and innovation, Absa.
Verushca Hunter, executive digital technologies and innovation, Absa.

The business value attached to AI and ML is heavily dependent on where organisations are on their digital transformation journey, and what the organisation is looking to achieve by using this emerging technology.

This was one of the points raised in a panel discussion on how to ensure that value is derived from AI and ML at ITWeb BI Summit 2024, at the Forum, Bryanston.

Panellists agreed that before organisations try to figure out how to get value, they need to understand what that value is or what the business objectives are that AI and ML will be used to try solve.

This is important, warned Ziaen Coetzee, a data strategist, who said because of the hype around AI, companies are triggered to quickly source and deploy the technology, and then try to find a use case.

This inevitably ends in failure and raises concern among business leaders.

“Value is linked to business outcomes,” said Coetzee. “Business value must be at the forefront. AI should be considered as part of the overall transition of a business; we need to have the building blocks in place first.”

Verushca Hunter, executive digital technologies and innovation at Absa, said: “We all look for short-term benefits, but it is actually a long-term play. I would say it’s important to have a clear strategy for at least the next 18 months; these are our focus areas, this is how we are going to do it – if you don’t do that, you will struggle.”

Mashoto Doubata, data product manager at Superbalist, agreed: “You must understand your business context and use cases that are linked to that context. Ultimately, you will have to show potential return on projects.”

Panellists added that successful use cases with AI are also about an organisation’s maturity level, something that is developed in stages, from initial data use to predictive analytics.

There is diagnostic value in AI, they said. The technology is used to review past events and their root causes.

Doubata said AI tools can be used to evaluate past sales, instances of increases and decreases, and root causes. “Now you are looking to predict future outcomes, what levers can be pulled to influence what the outcomes will be. How you view AI depends on where you are on your analytics journey.”

Vincent Maher, CEO of True I/O, focused on the role of AI in marketing as an important use case. “Marketing performance is one of the areas around value management or influence generated by marketing teams. They have their metrics, their performance metrics that are related to advertising, but this won’t make sense if you don’t tie it up to sales and data to perform marketing activities. Since ChatGPT, there has been an explosion of market technology businesses that use GenAI to gain further understanding of consumer behaviour. Google has been doing this for a long time and, frankly, Google and Meta dominate. AI can be used to compress time and perform rogue tasks that people are doing.”

Hunter said just-in-time training and the use of copilots as key trends are driven by the need for digital workforces, automation and to drive down costs.

The panel advised organisations to identify a small issue or challenge and empower a small group of dedicated professionals (who will not be put off by failure) to drive the project.