The £60,000 Question
Your full-time administrator is an illusion. Not the person, of course—they are very real, and likely very busy. The illusion is the role itself: a monolithic job title papering over a chaotic bundle of tasks ranging from the mission-critical to the mind-numbingly automatable. And for the privilege of maintaining this illusion, UK businesses are paying a steep price.
Let’s be brutally honest about the numbers. The average salary for an experienced Executive Assistant in London or Manchester hovers around £45,000, according to data from Hays and Reed. Factor in Employer's National Insurance, pension contributions, holiday pay, equipment, and the prorated cost of office space, and the 'fully loaded' cost to your business is north of £60,000 per annum. For one person.
Industry analysis suggests that person spends up to 40% of their time—two full days a week—on tasks that are now ripe for automation. Diary management, meeting transcription, expense report collation, information retrieval. This isn’t a critique of the individual; it's a condemnation of the anachronistic structure of the role. You are effectively paying a skilled professional a premium to perform tasks a £20 per month AI subscription could execute faster and more accurately.
This isn't an abstract economic debate. It’s the quiet, grinding inefficiency draining capital and focus from British scale-ups and established firms alike. It's the 'Admin Tax'—a hidden levy on productivity that we have accepted as a cost of doing business. It’s time for a radical reassessment.
The Great Decoupling: Why the 'Human Touch' is Overpriced
The traditional argument for a full-time, in-house administrator hinges on the amorphous concept of the 'human touch'. They are the cultural gatekeeper, the trusted confidante, the person who 'just knows' how the CEO likes their coffee. This is the controversial part: for the majority of administrative functions, the human touch is an expensive excuse for inefficient system design.
We have bundled two fundamentally different jobs into one role:
- Process Execution: High-volume, low-judgement, repetitive tasks. (Scheduling, data entry, formatting).
- Strategic Partnership: High-judgement, low-volume, high-trust tasks. (Pre-empting conflicts, managing sensitive relationships, high-stakes event planning).
The fallacy is believing one person is the optimal solution for both. By forcing a talented individual to spend half their time on automatable drudgery, you not only waste money but also dilute their capacity for the high-value work you desperately need. The real 'human touch' should be applied to strategy and relationships, not to finding a 30-minute slot in three executives' calendars.
The debate is no longer about simply hiring a person to fill a seat. The modern question for any leader is one of 'hiring vs building AI' capabilities. The answer, as it turns out, is a potent combination of both.
Deconstructing the Admin Stack
Instead of hiring a person, progressive companies are building an 'Admin-as-a-Service' (AaaS) stack. This isn't about replacing humans with robots; it's about augmenting human intellect with machine-speed execution. It consists of two layers: The AI layer and the Fractional Human layer.
Layer 1: 'Building' the AI Co-pilot
This is the foundation. For a fraction of the cost of even a month's salary, you can assemble a powerful AI toolkit to absorb the administrative load. The goal is to automate everything that follows a predictable pattern.
Consider this illustrative, yet entirely achievable, monthly AI stack for a small team:
- Intelligent Scheduling: Tools like Calendly or Reclaim.ai (£10/user/month) eliminate the endless email chains of finding a suitable meeting time. They sync with all parties' calendars and find the optimal slot automatically.
- Automated Transcription & Summarisation: A service like Otter.ai or the Fathom video assistant (£15/user/month) can join your Zoom or Teams calls, provide a full transcript, and generate an AI summary with action items moments after the call ends. The productivity gain here is immense—no more manual note-taking or debates over what was agreed.
- Inbox Triage & Management: Custom GPTs or tools like Superhuman (£25/user/month) can learn your email patterns, automatically sorting, prioritising, and even drafting replies to routine inquiries. This transforms the inbox from a reactive minefield into a managed command centre.
- Automated Expense Reporting: UK fintech darlings like Pleo (from £50/month for a team) provide smart company cards that automatically categorise expenses and chase receipts, eliminating the dreaded monthly expense scramble.
Total Monthly AI Spend: ~£100.
For roughly £1,200 a year, this AI layer can absorb an estimated 15-20 hours of administrative work per week across a small team. That’s the equivalent of a part-time employee, delivered with perfect recall and zero need for management.
Layer 2: 'Hiring' the Fractional Expert
AI is powerful, but it's not a panacea. It lacks judgment, proactivity, and the ability to navigate complex human dynamics. This is where the human element is re-introduced—not as a full-time, under-utilised employee, but as a high-calibre fractional resource.
Instead of a junior administrator, you engage a senior, UK-based Virtual Assistant (VA) or Fractional EA from a reputable provider like Time etc or Virtalent for a set number of hours. These aren't just task-doers; they are experienced professionals who have chosen a flexible career path.
A top-tier fractional EA might charge £35-£50 per hour. For 10 hours a week, this amounts to £1,400 - £2,000 per month.
What do they do in those 10 hours? Only the things the AI can't do:
- Manage the AI Stack: They are the 'human in the loop', training the AI, troubleshooting errors, and refining the automated workflows.
- Handle Exceptions: When the automated system fails—a last-minute flight cancellation, a double-booked VIP—they intervene with human ingenuity.
- Execute Complex Projects: Organising a multi-day investor roadshow, planning an off-site for 50 people, or managing a complex client onboarding.
- Act as a Strategic Partner: Because they are freed from mundane tasks, they have the mental bandwidth to be proactive, anticipate needs, and function as a true force-multiplier for the executive they support.
The Economics of the Hybrid Model: A Case Study
Let’s ground this in reality. Imagine a 25-person tech scale-up in Reading.
The Old Way: They hire a full-time Operations Manager / EA.
- Salary: £42,000
- Fully Loaded Cost (approx. 1.4x salary): ~£59,000 per year.
- Reality: The individual is constantly swamped, context-switching between chasing invoices, booking travel, and trying to manage the office lease renewal. High-value work suffers.
The AaaS (Admin-as-a-Service) Way:
- AI Stack: A robust suite of automation tools. Cost: ~£150/month = £1,800 per year.
- Fractional EA: A highly experienced professional engaged for 15 hours per week at £40/hour. Cost: £600/week = £31,200 per year.
- Total AaaS Cost: £33,000 per year.
The Result:
- Direct Cost Saving: £26,000 per year.
- Productivity Gain: The service level is objectively higher. An experienced professional is focusing exclusively on high-value tasks, while the AI layer provides 24/7 support for the low-value churn.
- Flexibility: The fractional model can be scaled up or down with a month's notice, providing agility that a full-time hire cannot.
At Creative Agent, we've seen this dynamic play out repeatedly. When we build custom AI workflows to automate a company's unique internal processes, the success of the project is directly correlated with the client’s willingness to re-imagine the human role alongside it. The most transformational results occur when a fractional expert is empowered to own and operate the new AI-driven system, elevating their contribution from task execution to system orchestration.
The Unbundling of the Job
The shift from a monolithic administrative role to a disaggregated 'Admin-as-a-Service' model is not a niche trend. It is a blueprint for the future of knowledge work.
We are witnessing the great unbundling of the traditional job. The rigid, 40-hour-a-week container of tasks, created for an industrial economy, is dissolving in the face of intelligent automation and a flexible, highly skilled freelance workforce. Every role, from finance to marketing to law, is a bundle of tasks with varying levels of complexity. The challenge for leaders is no longer 'who to hire', but 'how to build the most effective and capital-efficient stack of human and machine intelligence to get the work done'.
This requires a profound mental shift. It means moving from managing people to orchestrating systems. It means viewing AI not as a threat, but as the ultimate leverage for human talent. The companies that master this new operating model will not just be more efficient; they will attract the best people, who are increasingly unwilling to have their strategic talents squandered on work a machine can do. The £60,000 question isn't whether you can afford a full-time administrator, but whether you can afford not to embrace the future.
