The (Real) Challenge with the Diamond Organisation
Should we really be predicting the diamond is here to stay?
The World Cup in North America is just around the corner, and there’s nothing like a little reminiscing about past success — 1966 anyone?
It is also just over 14 years since Roy Hodgson replaced Fabio Capello as England manager and attempted to deploy what commentators called a “diamond formation”: narrow, technically elegant in theory, but ultimately lacking the balance needed to seriously compete. England exited the 2014 World Cup early and fared little better at Euro 2016.
Perhaps not all diamonds sparkle under the spotlight.
Now businesses appear to be embracing their own version of the diamond formation.
For decades, organisations have largely operated using pyramid structures: large numbers of junior employees at the base, fewer middle managers above them, and a narrow executive tier at the tip. The structure served two purposes simultaneously. It allocated labour efficiently, but it also created progression pathways through which firms identified, trained, and promoted future leadership.
AI may be about to disrupt both.
The logic behind the “diamond organisation” is now familiar. As AI automates more routine and administrative work, firms theoretically require fewer junior employees. Instead, they become weighted more heavily toward experienced specialists, orchestrators, and managers overseeing AI-enabled systems and workflows. The pyramid narrows at the bottom and expands through the middle. In theory, the organisation becomes a diamond.
But is that really what most industries will look like?
In some sectors — particularly technology, software, analytics, and certain forms of consulting — perhaps. Yet in many others, I remain unconvinced.
Retail still requires operational labour. Healthcare still depends on embodied human interaction. Education prioritises relationships, communication, and trust. Even in property, I struggle to envision armies of AI agents replacing frontline estate agents while management layers expand above them.
If anything, many organisations may end up looking less like diamonds and more like houses: flatter at the base, slightly broader through the middle, but still dependent on substantial human infrastructure.
There is also a practical constraint often overlooked in AI commentary. AI may lighten the workload, but there are only so many meaningful agentic tasks a single employee can realistically supervise, validate, maintain, report on, and action. Few organisations truly believe they are about to unleash “1000 cloned employees” through AI deployment alone.
Eventually, managerial attention becomes the bottleneck.
Whilst both the above points are reasonable, for me at least, the deeper issue with the diamond organisation is not operational. It is developmental.
Traditional pyramid structures were never simply labour systems. They were talent factories.
Entry-level roles provided organisations with more than low-cost execution capacity. They created pathways through which firms identified future managers, specialists, and leaders. The pyramid allowed organisations to observe performance over time, socialise individuals into organisational culture, and progressively develop institutional knowledge internally.
In many ways, it was an extended job interview for future leadership.
The diamond model weakens that mechanism.
Now don’t get me wrong. I do think firms will reduce graduate recruitment and junior hiring while AI absorbs routine work. Short-term efficiency will likely improve. Costs fall. Headcount shrinks. Contribution rises.
But eventually someone asks a more uncomfortable question:
Who replaces Michael when Michael moves upstairs?
At that point, firms may discover they have consumed the very talent pipeline that historically sustained the middle and upper layers of the organisation.
Football offers another useful analogy. Elite clubs can buy finished talent for a while, sometimes very successfully. But over time, many rediscover the strategic importance of academies and internal development systems. Imported talent is expensive, uncertain, and culturally volatile. Home-grown talent creates continuity, identity, and succession stability.
Organisations may discover the same thing.
The irony is that the traditional pyramid was not merely an inefficient labour structure waiting to be optimised away. It was also a long-term capability engine. Remove too much of the base, and firms may eventually realise they have hollowed out their own future middle.
And as every football fan knows, even the most elegant formation eventually collapses without a strong bench coming through behind it.