The Managerial Peloton is About to Break
There’s a concept in professional cycling known as the peloton. This main pack of riders creates a slipstream, reducing drag for those tucked inside by as much as 40%. It is the most efficient way to travel long distances. For decades, the corporation has operated on a similar principle. Founders and executives at the front, workers at the back, and a thick, protected layer of middle management tucked neatly in the middle, shielded from the headwinds of strategic risk and the turbulence of frontline execution. They draft. They report. They translate. And they are about to be exposed to a gale-force wind.
The narrative that AI will be a helpful ‘co-pilot’ for the knowledge worker is a comforting, incrementalist fiction. While true for some, it misses the tectonic shift. For the vast strata of middle management, AI is not a co-pilot; it is a replacement for the entire cockpit. The core functions that necessitated this managerial layer—information synthesis, delegation, and compliance monitoring—are precisely the tasks at which generative AI and sophisticated automation excel. The next 12-24 months will see a brutal, but necessary, culling of these roles. This isn’t a recessionary belt-tightening. It’s a permanent, structural change driven by a resurgent founder playbook, enabled by technology.
The £70,000 Report Generator is Obsolete
The fundamental value proposition of a significant portion of middle management has been information brokerage. They sit in meetings to gather data, which they then sanitise and repackage for senior leaders. They chase their teams for status updates, which they then consolidate into RAG-rated spreadsheets and slide decks. They are human routers, passing information up, down, and sideways.
This function is now a feature. An AI-powered dashboard connected to Salesforce, Jira, and Workday can generate a more accurate, real-time, and insightful report than a manager ever could. A custom-trained Large Language Model (LLM) can read every customer support ticket, sales call transcript, and internal Slack channel, and produce a daily sentiment analysis for the CEO that is richer and more honest than any summary filtered through three layers of management.
Consider the numbers. A mid-level Operations Manager or Project Manager in the UK earns, on average, a salary of £55,000 to £75,000. With National Insurance, pension contributions, and benefits, their total cost to a business easily exceeds £90,000. Industry analysis suggests that up to 60% of their time is spent on administrative coordination and reporting. That's a £54,000 annual cost just for information relay. A software solution that automates this costs a fraction of that and operates 24/7 with perfect accuracy.
Take a major UK retailer like Sainsbury’s or Tesco, with thousands of store and regional managers. Their primary role is ensuring operational consistency and reporting on performance metrics—stock levels, staff rotas, sales figures. Today, this requires a hierarchy. Tomorrow, a founder or CEO could ask a natural language interface: "Show me the five worst-performing stores for fresh produce spoilage this week, correlate it with staff absences, and suggest three optimised rota changes." The answer arrives in seconds, not in a PowerPoint deck three weeks after a regional meeting.
Automating the Compliance Chimera
Compliance is the second pillar propping up the managerial class, particularly in the UK’s heavily regulated sectors like finance and law. The argument has always been that you need human managers to ensure protocols are followed, boxes are ticked, and regulatory risks (from the FCA, ICO, or others) are mitigated.
This created a compliance chimera: part risk management, part bureaucratic inertia. Managers spend an inordinate amount of time enforcing internal rules, approving workflows, and documenting processes, not because it adds value, but because 'that's the procedure'.
AI doesn’t just automate this; it provides a demonstrably superior audit trail. An AI system can ensure every step of a new client onboarding process in a City law firm is completed and logged before the next stage is unlocked. It can flag suspicious transaction patterns in a fintech’s data stream in real-time, escalating only the genuine anomalies to a human expert, not a generalist manager. It can scan every outgoing email to ensure it meets GDPR requirements for handling personal data.
This moves compliance from a post-hoc, sample-based human audit to a pre-emptive, 100% consistent automated control. You no longer need a manager to police the process; the system itself is the process. The result is not only a vast reduction in the managerial overhead required for compliance but also a dramatic improvement in actual compliance.
The Resurgent Founder Playbook
This brings us to the core of the issue: the founder playbook. The archetypal founder—from a Shoreditch tech start-up to a Wrexham manufacturing challenger—craves speed, agility, and a direct line of sight from vision to execution. Bloated hierarchies are anathema to this. Middle management creates communication latency, dilutes strategic intent, and fosters a culture of risk-aversion and consensus-building.
For years, founders tolerated this bloat as a necessary evil of scaling. Once a company grew beyond about 150 people—Dunbar's number—direct communication became impossible, and structure was needed. Middle management was that structure. A 20th-century solution to a 20th-century communication problem.
AI offers a 21st-century solution. It allows a founder or a lean executive team to maintain a direct connection to the entire organisation. It creates a 'synthetic nervous system' for the business.
At Creative Agent, we've seen this shift in demand firsthand. Founders aren't commissioning us to build tools to help their managers write better reports. They are commissioning systems to replace the functions of management altogether: intelligent resource allocation models, automated performance tracking that provides direct feedback to employees, and cross-functional knowledge bases that eliminate the need for a human information silo.
This is a conscious decoupling of growth from headcount. A UK scale-up can now aim to reach £100 million in revenue with 200 employees, not 800. This is a profound shift in the economics of business building, and middle management is the primary cost being designed out of the system.
Contrarian Take: This is a Mercy Killing
The polite discourse around AI and jobs is that we must manage a 'just transition' for displaced workers. This framing is flawed. The culling of the managerial class is not the unfortunate collateral damage of technological progress; it is the overdue and necessary correction of a decades-long organisational pathology.
Middle management, in its current form, is often a soul-crushing black hole of talent. It takes bright, ambitious people and turns them into PowerPoint jockeys and meeting facilitators. It promotes individuals to their level of incompetence, a principle that has never been more relevant. A recent poll by the Chartered Management Institute found that 82% of new managers have had no formal management training. We have created a vast class of 'accidental managers' who are neither good leaders nor effective individual contributors.
AI will not destroy value by replacing them; it will unlock it. It will liberate capital to be invested in R&D, customer service, and rewarding genuine top performers. More importantly, it will liberate the talent currently trapped in these roles. The brilliant engineer who was promoted to a non-coding management position can go back to being a brilliant engineer—and be compensated as one. The sharp financial analyst who became a team leader can become a principal analyst, a domain expert of immense value.
This isn't about creating a flat, anarchic organisation. It's about replacing a hierarchy of people with a hierarchy of information. Leadership, mentorship, and strategic decision-making remain profoundly human endeavours. But the administrative, supervisory scaffolding that we have long mistaken for management is about to be dismantled. It’s not a tragedy; it’s a mercy killing for a role that had long outlived its purpose.
The Unbundling of Management
The next twelve months will not see a CEO stand up and announce, "We are firing all our middle managers." The change will be more subtle, executed through a thousand individual software procurements, process re-engineering projects, and 'efficiency' drives. It will be a slow-motion demolition.
Roles will be consolidated. When a manager leaves, they won't be replaced. New 'Head of...' titles will appear, but they will oversee automated systems and a small cadre of elite experts, not armies of report-writers. The very concept of a manager as a people-supervisor for a team of eight will seem archaic.
We are witnessing the great unbundling of the managerial role. The tasks are being stripped away, automated, and optimised. What remains is the human core: coaching, creativity, and complex judgment. Anyone whose job does not revolve around these three things is standing on a trapdoor. And the lever is about to be pulled.
