Productivity Before Headcount

For CPOs, the productivity conversation is becoming harder to separate from workforce strategy. As executive teams look for greater efficiency and AI capability accelerates, People leaders are increasingly being drawn into decisions about where work should sit, which capabilities remain critical and whether existing workforce structures are still fit for purpose.

That conversation is happening against a challenging economic backdrop. The Productivity Commission’s September update showed Australian labour productivity was flat in the June quarter and 0.2% lower over the year to June, reinforcing the pressure on organisations to find more effective ways of working.

Are we redesigning work before reducing it?

As organisations look to AI for productivity gains, one of the more important questions is becoming one of sequencing. Do you redesign the work first and determine what technology can genuinely absorb, or make workforce decisions in anticipation of efficiencies that have not yet been proven?

AHRI recently explored exactly that scenario, asking how HR should respond when an executive team proposes cutting 15% of a team before an agentic AI tool has been appropriately stress-tested. The example is hypothetical, but the tension is very real: workforce decisions made too early can lock organisations into assumptions about technology, capability and capacity before there is enough evidence to support them.

For CPOs, this creates a broader strategic question. Productivity should not simply mean doing the same work with fewer people. The better conversation is about how work itself should change: which processes can disappear, where automation genuinely adds value, how roles should be redesigned and where human judgement remains essential.

Headcount may ultimately change as a result. But it should not necessarily be the starting point.

AI can create work as well as remove it

There is another part of the productivity equation worth paying attention to. AI does not always remove work cleanly; in some cases, it shifts work into reviewing, correcting and supervising AI-generated output.

Recent AHRI coverage of research involving 6,000 knowledge workers across Australia, the US and the UK found that employees reported spending 6.4 hours per week supervising AI tools. Australian respondents reported 6.5 hours a week on this work, while also saying AI completed the equivalent of around 10 hours of work for them.

That does not diminish the opportunity AI presents. It does suggest that productivity gains need to be measured in practice, rather than assumed at implementation.

There is also a longer-term capability question. As AI absorbs more routine work, organisations will need to be deliberate about where future judgement, experience and capability are developed.

The CPO question is getting bigger

For CPOs, the opportunity is to move the conversation beyond “how many roles could AI replace?” and towards a more useful question:

What should this organisation look like if we redesign the work properly?

That opens a much richer discussion about capability, structure, spans and layers, talent pipelines, decision rights and where human contribution creates the greatest value.

With productivity pressure unlikely to ease, this feels like an important discussion for CPOs heading into 2027.



Where is the balance landing in your organisation: work redesign first, or workforce reduction first?