A New Kind of Company Is Emerging

At a recent HyperscaleCEO cohort session, Matthew Scullion, CEO of Matillion and Hyperscale’s Entrepreneur in Residence, spoke about a shift he is seeing first-hand across the market. 

Matthew described a new kind of company that was emerging: AI-native organisations reaching millions in revenue with remarkably small teams, often between 10 to 50 people. Many are lightly funded or even bootstrapped, yet their revenue per employee far exceeds anything we have historically associated with SaaS. In these companies, ambition no longer scales linearly with headcount.

This isn’t a future prediction, it’s already happening. It forces a difficult question for leadership teams: why do some companies reach 50m in revenue with 30 people while others struggle to reach 10m with twice the headcount?

One explanation lies in the 10 / 80 / 10 rule. Roughly ten percent of people in an organisation truly reinvent how work happens. Another ten percent resist or disengage. The remaining eighty percent sit in the middle, making marginal gains while keeping their existing workflows largely intact. What determines outcomes isn’t what leaders do about the resisters, it’s what they do about the middle eighty percent. This is where most companies stall.

The reason this happens is not a lack of talent or effort. It is structural and behavioural. The top ten percent redesign the work itself: They collapse roles, eliminate handoffs and fundamentally redefine what one person can accomplish. Job descriptions lose their meaning. Boundaries blur. The work reorganises around outcomes, not functions. In contrast, the middle eighty percent adopt AI tools but keep the same roles, the same interfaces and the same decision rights. In essence, AI is layered on top of old structures rather than used to dismantle them.

Here's the painful reality for us all: If you are working the same way you were a year ago, that should be a warning sign. AI adoption without workflow redesign creates the illusion of progress while locking in yesterday’s operating model.

When reinvention actually happens, it looks very different from productivity optimisation. In product teams, a single individual may synthesise customer input and design the user experience in days rather than weeks because there are no handoffs between research, strategy, and design. In engineering, humans increasingly orchestrate intelligent agents rather than writing most of the code themselves, shifting the role from implementation to direction and massively increasing output per person.

In sales, individuals can run deep, tailored, end-to-end account strategies from a single laptop, with research, outreach, positioning and follow-up all supported by AI. The emphasis is not on tools but on scope. Fewer people are doing more because the work itself has been redesigned.

One of my favourite coaching topics as many of you know is the importance of talent density in an organisation, and it is in this world of business reinvention where talent density becomes super critical. When every team member is highly capable and able to operate with autonomy, the effects of collapsing roles and eliminating handoffs compound. A small, dense team of exceptional talent can accomplish what previously required dozens of people. In contrast, adding AI to low-density teams produces only incremental gains and can even highlight weaknesses. High talent density is no longer optional; it is a force multiplier for companies looking to thrive in the AI era.

Most leadership teams, in my experience, genuinely believe they are invested in AI. They have brought in tools, run training sessions and told their organization to fully embrace this brave new world. Yet a year later, workflows look almost identical. The real question is not whether you are using AI. The real question is whether you are willing to redesign the work itself, including roles, boundaries, and ownership, or whether you are simply optimising the structures you already have. 

What gets measured reveals the truth: Revenue/profit per employee becomes a signal of whether work has actually changed, or merely been automated at the edges. The number of handoffs eliminated shows whether complexity has been removed or preserved. The scope of individual output reveals whether one person can now do what previously required three, or whether the old model is still intact under a new layer of technology.

The companies that will dominate the next decade are not waiting for permission, consensus or the perfect toolkit. They are collapsing roles, eliminating handoffs and redefining what one person can accomplish. The rest are getting better at a way of working that may not exist in 12 months time.

So the question is simple, but uncomfortable: which kind of company are you building?

As you reflect on this, here are three questions every CEO should be asking themselves:

• Where in my organisation are we adding AI on top of existing workflows instead of redesigning the work itself? 

  
  

• Which roles could radically expand in scope if we removed handoffs and decision bottlenecks? 

  
  

• If we froze headcount today, what would we have to change about how work happens in order to double output?

Those answers will tell you far more about your future than any AI roadmap ever will.

Photo by Cash Macanaya on Unsplash