Why the AI KPI list needs rewriting – and productivity bumped to the top

Yomi Tejumola, founder and CEO of Algomarketing, says that shifting AI adoption from an efficiency mindset to a productivity one could deliver far bigger gains and avoid unnecessary job losses

So far, the vast majority of the discussion around AI has centred on the efficiencies it can drive within organisations: automating tasks, reducing headcount, streamlining workflows. It’s the same playbook businesses used for every wave of technology before this one and, as a KPI, it makes sense. Cost and time reduction feels tangible and immediate, while fitting neatly within the quarterly targets most leaders are under pressure to hit.

More often than not, however, efficiency – in other words, doing the same work with fewer resources – turns out to be a euphemism for shrinking input and cutting jobs. A recent British Standards Institution survey found that 41% of business leaders across seven countries said the technology was allowing them to reduce employee headcount.

The problem with cutting too deep in the pursuit of cost savings is that it breeds fragility, and you lose the human context that makes AI valuable. Specifically, this may mean siloed automations that don’t connect across teams, resistance from employees who see AI as a threat, or lost innovation, because you remove the people who understand nuance and customer behaviour.

Why productivity is a better KPI

The alternative approach is for management teams to make a subtle mindset shift, looking at the technology as a driver of productivity over efficiency. Productivity measures the human–machine partnership – how teams think faster, learn faster, and deliver better outcomes – and recognises that people plus AI outperform either alone. 

It’s also the moral differentiator. Productivity-led AI, where we create more value with the same resources, expands capacity, creativity, innovation and output. In other words, it creates much more interesting, human-centered work, meaning less repetition, more problem-solving and more decision-making. Equipping workers with the capabilities to do more is an approach that’s less likely to see roles automated and people replaced, too. 

Think of it in terms of reporting. AI for efficiency might automate reporting, while AI for productivity reimagines what you report, why, and how fast you can act on those insights. One replaces; the other amplifies.

Overall, the latter is how you build organisations that grow capability, not fear, and it’s already proving effective. Last year, Goldman Sachs Research said that the early signals of future productivity gains from AI were ‘very, very positive’. Joseph Briggs, who co-leads the Global Economics team in Goldman Sachs Research, said specifically: “Some of the academic literature and economic studies that have looked at the increase in productivity that we’ve seen following AI adoption, in a few specific cases, supports our view that large productivity gains are possible. The average increase in productivity is about 25%.” 

How businesses can measure and communicate productivity gains

But to turn those early signals into something meaningful, businesses need a clear way to measure and communicate where productivity gains are actually showing up.

My advice here is to start by shifting the questions you ask:

• How many hours of manual effort were reallocated to higher-value work?
• How has cycle time improved on projects, not just tasks?
• Are employees spending more time on strategic or creative work?
• Has our innovation curve increased and are more people bringing ideas to the table?
• What new initiatives or clients have been won through increased capacity?
• Are teams adopting and utilising new AI tools? And are these driving business impact?

Then look at how AI is changing the way teams work, without reducing headcount. Many organisations are already seeing gains as AI automates the “boring middle” – for instance, routine admin or reporting that previously slowed teams down. Others are using it to uncover surface insights rather than just raw data, helping people act faster and make better decisions. Or perhaps it’s helping accelerate learning, whereby new hires are getting contextual coaching from AI systems, reducing ramp-up time without replacing mentorship.

Whatever the use case, the next step is to communicate this in human stories, not just metrics. Look to share examples of teams achieving things that weren’t possible before – a faster product launch, a campaign with hundreds of multi-variate tests, a new insight surfaced through AI co-analysis – and focus on the broader impact this creates, from more confident decision-making to teams pushing out ideas they never previously had the capacity to pursue.

Smarter, not smaller

Ultimately, the real opportunity here isn’t about saving a few pounds, it’s unlocking what’s possible. When teams reclaim time and redirect it into higher-value work, you get more personalised delivery, faster cycles of experimentation and sharper insights. 

Efficiency becomes momentum, enabling teams to do in days what once took weeks, and keeping them ahead of competitors still stuck in cost-cutting mode. 

These teams aren’t smaller; they’re smarter, with the space to focus on the work that genuinely moves the business forwards.

About Yomi Tejumola 

Yomi Tejumola is an award-winning entrepreneur and visionary technologist on a mission to reshape how enterprise marketing teams operate. As the Founder and CEO of Algomarketing, he leads a global talent solutions firm built to embed AI-enabled marketers directly into modern marketing organisations.