The Shamrock Organisation

Something like fourteen years ago (crikey) I wrote about the concept of the shamrock organisation, and it’s something I’ve found myself coming back to, notably in my thinking around agentic AI. In the late 80s Charles Handy, in his book The Age of Unreason, used the shamrock metaphor (he was Irish of course) to describe organisations whose resourcing would be comprised of three distinct parts: core staff and FTEs who are essential to continuity and organisational knowledge; a ‘contractual fringe’ of specialists that operate with flexibility within defined guidelines; and a flexible work-force of part-time, temporary or seasonal roles. All those years ago I was using Handy’s metaphor to describe a need for increasingly flexible ways of working and approaches to resourcing.

In the age of agentic AI, where agents will increasingly be working alongside humans to complete processes and tasks, we can reclassify the shamrock metaphor as a way of articulating the need for organisations to balance resourcing in workforce design across three main areas:

• Full-time employees: still very much the core of resourcing, and essential to capability, culture, continuity, experience and sensibility. FTEs may well have deep institutional knowledge, strategic or executional understanding that is hard to replicate, good alignment with company values, good commitment, and key to the growth and development of the business and each other.
• Freelance and contract staff: external to the organisation but useful for specific needs or projects, fluctuating resource requirements, enabling the organisation to be flexible but also access specialist knowledge and expertise.
• AI agents: In January Sam Altman said that this year AI agents would be integrated into the workforce and would ‘materially change the output of companies’ and so (like it or not) they will become a growing part of the workforce, initially good for repetitive tasks, data processing, predictive analytics, and enhancing productivity, scalable and potentially always on, potentially freeing up people for higher order tasks.

Another way of thinking about this ‘Shamrock organisation’ is that instead of thinking in rigid roles, organisations should structure their workforce like an ecosystem, with three dynamic layers:

• Core (FTEs): The backbone of the organisation, responsible for leadership, strategy, company culture, and human-driven tasks and oversight.
• Periphery (freelancers & contractors): A fluid network of specialists who inject innovation and specialist knowledge, handle seasonal demands, and support agility.
• Exponential (AI & automation): a scalable intelligence layer that performs repetitive, data-driven, and predictive tasks to enhance human capabilities.

If this hybrid model is going to become a reality, I think it’s important that we take a deliberate, intentional approach to design this balance so that it works for the benefit of people as well as the organisation – a way that maximises efficiency and agility, as well as employee well-being. There are, perhaps, some key principles that are useful here:

• People-centric design: it’s going to become increasingly important that the needs of people are baked into workforce design to ensure that AI and automation can enhance human work and that efficiency does not come at the expense of an ethical and inclusive approach.
• Task-based resourcing: an appreciation of the nature of the work (stable vs dynamic, consistent vs variable, high-risk vs low risk, strategic vs executional, creative vs routine) will enable better matching of different types of resourcing.
• Workforce agility: building a flexible model that can scale up and down based on business cycles whilst still allowing for appropriate protections for people
• Continuous optimisation: as markets, business and employee needs change, there will be a requirement for regular review, adjustments, and adaptability.

Clearly there’s going to be a need for workforce mapping and role categorisation to break down work functions in order to map tasks to resource types in the best way. And there will also need to be an intentional approach to AI/Human collaboration, a focus on AI literacy, and also plenty of transparency to ensure trust.

There will also need to be ongoing flexibility, a new level of agility to workforce planning, and smarter approaches to tracking both workforce productivity but also employee engagement. It strikes me that this will need to go far beyond traditional workforce planning and become a self-adjusting, dynamic system that evolves with the needs of both the business and employees.

The ‘human-in-the-loop’ approach that I described in my agentic presentation, and an awareness of context and risk in decision-making is going to be needed to ensure that AI agents can augment human capability in the right way. Broadly speaking, there’s likely to be three fundamental roles that AI agents can play:

• AI as the assistant: automating repetitive tasks
• AI as the collaborator: providing recommendations, insights, options and co-creating with humans
• AI as the decision partner: using predictive analytics, scenario planning and simulations to guide high-level business strategy

The implications of this are huge. But the biggest risk is that it happens in a way that destroys rather than builds employee trust, and that employee engagement is eroded even further than it is now (from an already low base). We have a choice to do this the right way, or in a way that creates a myriad of unforeseen effects and consequences.

I write a weekly Substack on the latest in AI, digital strategy and trends, transformation insights and the best links on the internet. To join our community of over 10,000 subscribers you can sign up to that here.

To get posts like this delivered straight to your inbox, drop your email into the box below

Photo by Thorsten Hack on Unsplash