Agentrification: Where Work and Value Move

When agents take on tasks, work does not simply disappear. It is rearranged. Some activity moves to machines, some is created, and some changes hands. The people and organisations who understand where it goes will be better placed than those who simply count how many jobs are "at risk".

I use the word agentrification for this process: the progressive takeover of activity by agents inside a role, a firm or an industry, and the shifts in value that follow. Like gentrification, from which the word borrows its shape, it does not happen all at once, and it changes who benefits from a place long before the change is visible on the surface.

This essay sets out how agentrification works at three levels, the hard question it raises about how people learn, and how leaders can trace where value is moving in their own sector.

Inside the role

Most jobs are bundles of tasks. Agents rarely replace a whole job at once. They take over tasks within it, starting with those that are frequent, well defined and digital.

As that happens, roles move in one of two directions.

Some roles narrow. When an agent handles the routine cases, the person is left with supervision and exceptions: checking samples, handling the unusual, approving what the agent cannot. The work becomes less varied and, for some, less satisfying, even as it becomes more responsible.

Other roles expand. One person directing several agents can do work that once required a team. An analyst becomes the manager of an analytical process; a relationship manager serves more clients with better preparation. These roles reward judgement, direction and accountability more than execution.

The same job title can go either way, depending on how the organisation redesigns the work. That is a management choice, not a technological inevitability.

The apprenticeship problem

The tasks agents take over first are often the tasks junior people have always done: first drafts, data gathering, reconciliations, routine cases. Those tasks were never only about output. They were how people learned their trade, built judgement and earned trust.

If entry-level work is automated without replacement, organisations save money today and lose their future leaders tomorrow. A firm that no longer trains juniors will, in ten years, have few seniors who understand the work from the ground up, and no one able to supervise the agents with real expertise.

This is the hardest question agentrification raises, and few organisations are yet asking it. Answers will include deliberate rotations through supervisory work, simulated cases, closer apprenticeship to senior staff, and treating the development of judgement as an explicit cost of doing business rather than a by-product of junior work.

Inside the firm

At the level of the firm, agentrification changes the shape of the organisation. Layers that existed to coordinate, check and pass information upward thin out when agents do that work. Spans of control widen. The cost base shifts from people towards technology, data and assurance.

It also changes what the firm must be good at. When execution becomes cheap, competitive advantage moves to the things agents cannot easily replicate: proprietary data, trusted relationships, regulatory standing, brand and the judgement to direct agents well.

Across the industry: where value moves

The same forces act on whole markets. When the cost of skilled labour stops being a barrier, advantages built on it erode. Margins tend to move:

• towards those who own the data an agent needs, the customer relationship it serves, or the trusted brand that makes people willing to accept its decisions;
• away from those whose main advantage was the cost or scarcity of skilled people performing routine work.

This is already visible in professional services, where routine drafting and research are becoming cheaper, and value is moving to firms with distinctive expertise, data or relationships. It will reach other sectors as agents become capable in their work.

Tracing value in your own sector

Leaders can trace agentrification in their own organisation with a simple method.

1. List the work, not the jobs. Break roles into tasks and estimate how much time each takes.
2. Mark what agents can do now, and soon. Be specific: which tasks, with what reliability, under what supervision.
3. Follow the cost pools. Identify where the organisation's costs sit today, and which will shrink as tasks move.
4. Ask who captures the saving. Will it stay with the organisation, pass to customers through lower prices, or flow to whoever owns the data and the platforms?
5. Identify what becomes scarce. When execution is cheap, what will customers still pay a premium for?

The answers indicate which capabilities to build, which to protect, and which partnerships or acquisitions might matter.

Managing the human transition with candour

The way organisations talk about this matters. Euphemism damages trust. Telling people their roles will simply be "augmented" when the plan is to reduce headcount is quickly found out, and it makes every later conversation harder.

Candour does not mean announcing redundancies before decisions are made. It means being clear about which kinds of work will change, what the organisation will do to help people move into the roles that grow, and how decisions will be made. Organisations that are honest about agentrification are more likely to keep the people they most need.

Questions for leaders

1. Which tasks in our organisation will agents take over first, and which roles hold them?
2. How will people learn the judgement that junior work used to teach?
3. Where does value in our industry move when skilled labour is no longer the constraint?
4. What do we own that agents need, and what do we depend on others for?
5. Are we describing the change to our people honestly?

What follows

If agents rearrange work and value, every agent should be able to show what it contributes. Part IV introduces the value ledger: a discipline for making each agent earn its place in the accounts.

This essay introduces Part III of the forthcoming book The Age of Agency.

Views are the author's own. The frameworks described are applied methodologies of Kiros Intelligence Group, not regulatory standards.