Is the Pyramid Still Fit for Purpose?

The partnership model in professional services is built on an elegant logic. At the top, a small group of partners setting direction and managing client relationships. Below them, a middle layer organising and supervising the work. And at the base, a broad intake of young professionals carrying out routine work, learning through doing, and progressing over time. That broad base is more than a training ground. It is also the economic engine of the model: juniors generate more value than they cost, and in doing so finance the partner layer above them.

AI is now eroding that base. Not suddenly, but structurally.

What the numbers already show

The signals are concrete. Randstad reports that the number of entry-level vacancies in the Netherlands has fallen by 30% since January 2024. CBS data shows that in the third quarter of 2025, more than 13% fewer 15- to 24-year-olds were working in roles heavily dependent on knowledge work compared to the end of 2022. In those same roles, the number of open vacancies in Q2 2025 was nearly 25% lower than at the end of 2022. In roles outside this category, the decline remained below 10%.

Researchers are clear on one point: the decline cannot be explained by economic headwinds. Sectors such as business services actually grew in added value. Yet the number of young workers there fell significantly. The explanation is simpler, and more uncomfortable: companies are hiring fewer entry-level staff because AI is taking over a growing share of the work those staff were recruited to do.

PwC illustrates this in its own annual report. The intake of new employees fell by 23% in a single year: from 1,125 to 870. The year before, the figure was 1,356. The decline is explicitly at junior level, confirmed by PwC's own CFO. Among the causes cited by chief executive Agnes Koops-Aukes: the deployment of AI and the anticipation of its impact.

The problem is not in the layer that is disappearing

Until now, the debate has focused primarily on whether AI is costing jobs. That is the wrong question. The more relevant question is what happens to the model when the base becomes structurally narrower.

Fewer juniors means less progression. Less progression means a thinner pipeline of future seniors. And fewer seniors means, over time, fewer partners, and fewer people along the way financing the partner layer above them. The pyramid that has been discussed for a decade is now beginning to take on a different shape.

What I observe daily in my conversations with executives and partners is that every significant shift follows a recognisable pattern. This is no different. First there is unease and considerable noise. Then, broadly, three groups emerge. The first is curious and actively experimenting. The second is watching and waiting. The third keeps its distance and has limited visibility of the impact. All three exist simultaneously, in almost every organisation today.

At the same time, AI has risen firmly to the top of the agenda. Yet the number of organisations that translate this into a fundamentally different way of working remains limited.

The question that is rarely asked

Many firms are investing in AI tooling. They are right to do so. But the strategic question that comes with it is rarely asked openly in boardrooms: if we are hiring fewer entry-level staff because AI is taking over their work, how do we develop the next generation of seniors? And who will finance today's partners in ten years' time?

The model on which professional partnerships are built assumes a broad intake base. That base was never simply a cost item. It was also a nursery for knowledge, culture and continuity.

There are exceptions. Firms such as PwC and Deloitte are not only investing in tooling but also adjusting their intake strategies and experimenting with different team structures. In the technology sector, companies such as Klarna and Shopify are actively deploying AI to make operational teams smaller and more specialist. But across most of the market, exploration has yet to become redesign.

What is already visible is that the intake is changing. The demand has shifted from generalists to specialists. The number of junior positions is declining, and the threshold for the intake that does remain is considerably higher. The question that becomes ever more relevant: what is the added value of an entry-level hire when part of the work is carried out by AI?

What this asks of leaders

The pyramid is changing shape. Whether that becomes a problem or an opportunity for deliberate redesign depends on whether leaders are willing to ask this question now. Not as an HR matter, but as a strategic choice about the future of the model itself.

AI is taking over work, but it is also creating capacity. Particularly in the lower and middle layers, room is opening up. How that room is used will be a strategic choice. That places greater pressure at the front of the process: leadership and HR must determine, with more precision than ever, what talent they bring in and why. Development remains essential, but the timeframe and context in which it takes place are shifting.

For leaders, this means: moving faster, taking a position, and accepting that not everything can be predicted. This is not a linear change, but a phase of experimentation and exploration. Looking back at earlier technological transitions, the same pattern emerges. There was uncertainty and resistance then too. And in retrospect, those transitions did not weaken economies but strengthened them.

A recent conversation with a young, highly tech-savvy professional made this personally clear to me. His understanding and application of AI were well ahead of what I encounter in many boardrooms. That kind of insight should not remain on the sidelines. It needs to be actively brought inside.

Perhaps that is also my role. Not only in connecting executives with one another, but in bringing together different generations and perspectives. To challenge leaders not only on leadership, but on substance, and on what they do not yet know.

At Odgers, we believe the best advisers keep moving, keep questioning, and bring others along on that journey. Do you recognise these challenges, or would you like to explore the topic of leadership in complex circumstances? Please get in touch with Rosanne.