The Inside View

With Duncan McNeill

The Inside View goes beyond the headlines. It's where we share candid insights from our client meetings — the trends they're navigating, the talent challenges they're facing, and what it all means for the market.

Inside View

Why FMCG's Biggest AI Returns Aren't Coming From Head Office

Most of the AI conversation I hear in finance circles is about head office. Who has rolled out which tool, how many licences have been bought, whether analysts will still exist in five years. It's a fair conversation. But after a catch-up this week with a finance leader at a large Australian food producer, I'm convinced it's the wrong place to look if you want to see where AI is actually paying its way in FMCG.

The real returns are in operations. And that's not new.

 

Head Office: High Adoption, Hard-to-Measure Return

This business rolled out a generative AI assistant to its corporate staff around two years ago. Adoption has been strong. Accountants use it to work through technical questions, planners use it to speed up analysis, and most people would say it makes their day easier.

But ask what it's worth to the bottom line and the answer gets vague. Six months ago they capped licence spend per user after costs ran ahead of value. That's a pattern I'm hearing from a lot of clients:

  • People use AI to do parts of their existing job faster, but it's unclear whether the time saved goes into higher-value work or simply disappears.

  • A fair share of usage is personal rather than productive.

  • It's excellent at crunching data and producing a first draft of analysis, but it still makes assumptions no finance professional would make. Someone with real judgement has to check the output.

None of that means head office AI isn't useful. It means the gains are incremental and hard to quantify.

 

Operations: Measurable, Material, and Long Established

Contrast that with what the same business is doing in its supply chain. It's trialling cameras, lasers and sensors in its growers' sheds, with software that gives early visibility of performance long before the product reaches the processing plant. Vets get out sooner, problems are caught earlier and yield improves. The finance team's view was blunt: the return is far higher than anything head office use could deliver.

Another FMCG client of ours has just opened a manufacturing facility built around AI-driven automation, from production lines through to logistics. They're understandably tight-lipped on the numbers, but you don't commit that level of capital without confidence in the payback.

Here's the point that often gets lost: this isn't a ChatGPT-era story. Machine vision for quality control, predictive maintenance on plant equipment, demand forecasting and automated warehousing have been part of FMCG operations for years. Manufacturers were applying machine learning to real physical problems well before generative AI made AI a boardroom talking point. What's new is the label and the hype, not the operational use.

 

Why Operations Gets the Bigger Payoff

  • The problems are physical and measurable. Yield, waste, downtime and throughput have clear dollar values. Improvements show up directly in margin.

  • Scale magnifies small gains. A 1% improvement across a high-volume, thin-margin operation is worth far more than a few hours saved on a month-end report.

  • The investment case is built upfront. Operational AI usually needs significant capital, so it gets a proper business case. Head office tools are often bought by the licence and justified later.

  • Margin pressure demands it. With input costs rising, new competitor capacity and retailers pushing hard, FMCG businesses need hard savings, not soft productivity.

 

What This Means for Finance Talent

This is where it gets interesting from a recruitment point of view.

Manufacturing and supply chain finance professionals have always been among the hardest people to find in commercial finance. If the biggest value from AI is being created on the factory floor and in the supply chain, the finance people who understand those operations become even more valuable. Businesses need people who can build the investment case, track whether the returns arrive and challenge the numbers.

At the same time, there's a real dilemma in head office. If AI absorbs the work traditionally done by analysts, where will the next generation of experienced reviewers come from? You can't develop finance judgement without doing the groundwork first. Businesses that cut their junior pipeline too hard now may find themselves short of capable senior finance people in ten years.

 

The Takeaway

If you're in FMCG and judging your AI progress by head office licence numbers, you're probably looking in the wrong place. The businesses getting real value are those applying it where they always have: to the physical, measurable problems in operations. Finance leaders who understand that, and can link AI investment to hard operational outcomes, will be in high demand.

I'd be interested to hear from others in FMCG. Is your biggest AI return coming from operations or corporate?

Share: