Sandton:

Before Investing in More Finance AI, Measure Where the Time Is Actually Going

Finance leaders across South Africa are approving new AI tools, Microsoft Copilot licences and automation projects faster than they are measuring what those tools actually change. The intention is sound. The evidence behind the decision is often missing.

How much of your finance team’s time is spent producing the numbers rather than understanding what those numbers are telling the business?

Most finance leaders cannot answer that question with a figure. They can describe the month-end pressure, the reconciliation delays and the late nights before a board pack is due. What they usually cannot say is how many hours, across how many employees, are consumed by work that adds no analytical value at all.

That gap matters, because it is difficult to improve what has never been measured.

EY South Africa’s research into CFO priorities identifies four recurring pain points: establishing dependable data, maintaining a connected view of performance, improving reporting efficiency and freeing finance professionals for higher-value work.

EY makes a further point that organisations often overlook — solving these problems is rarely a matter of implementing one new platform. The improvement usually depends on the underlying process, not the technology layered on top of it.

“This is a pattern College Africa Group encounters when discussing workplace productivity and finance processes with organisations.”

A team becomes highly efficient at running an inefficient process. Data is exported, spreadsheets are updated, figures are reconciled by hand, and the report eventually balances. It has worked every month for years, which is exactly why it goes unexamined.

The risk is that organisations respond to this pressure by buying technology before they understand the workflow it is meant to fix. A new AI licence can accelerate a process without making it more valuable, particularly when nobody has established what the process should actually be producing.

“A business does not get a return from owning technology. It gets a return when technology improves the way work gets done,” says Arnold Muscat, Managing Director of College Africa Group. “Before an organisation can claim that improvement, it needs a baseline — otherwise it is simply assuming the tool is working.”

A measurable baseline does not need to be complicated. It typically covers the employee hours a recurring report consumes, the amount of manual correction and reconciliation involved, the frequency of rework, and the time between month-end and management receiving usable information.

Once that baseline exists, an organisation can test whether a change — stronger Excel capability, Power Query, Microsoft 365 collaboration, Copilot or AI-assisted analysis — actually reduces the work, rather than assuming it does.

College Africa Group’s methodology treats this as a sequence rather than a technology decision: Source Data ? Process ? Required Output ? Skills & Technology ? AI ? Human Validation ? Business Insight. For AI specifically, the discipline narrows further — Source Data ? Process ? Required Output ? AI ? Validate Against the Source.

AI can accelerate analysis, draft commentary or summarise information, but a finance professional remains responsible for verifying that the output reflects what the source data actually shows.

This is also why stronger workplace skills matter more, not less, as AI adoption grows.

An experienced Excel user is better positioned to notice when an AI-generated formula does not make sense. A finance professional who understands the underlying process can distinguish a genuine productivity gain from a faster version of the same inefficiency.

The outcome organisations should be measuring is not whether AI has been deployed. It is whether skilled finance employees now have more time to analyse margins, investigate unusual movements, assess risk and support management decisions — the work a report alone cannot do.

The College Africa Group Perspective

Before approving another finance tool, establish what the current process is actually costing in time, rework and delay. College Africa Group’s AI Workflow Discovery helps organisations build that baseline and determine whether the right response is process redesign, stronger Excel and Microsoft 365 skills, automation, AI, or a combination of these.

Explore AI Workflow Discovery: AI Workflow Discovery South Africa

Speak to College Africa Group: WhatsApp or call 083 778 4903.