The Always-on Agency: How to Survive the Shift to Intelligence-Native Organizations

We've spent three years asking whether AI will take our jobs.  Recently I have been wondering if this is the wrong question entirely. The one that keeps me up at night is this: are the companies we work for evolving quickly enough for an AI-native world – one where intelligence moves continuously and organizations never sleep?

The Model That Is Breaking – Death to the Pyramid

It was in the 1840s that the traditional business hierarchy in the shape of a pyramid started to appear - designed to bring military organizational thinking and discipline to the companies building the railroads.  This structure has certainly stood the test of time with many organizations still using it today. 

However, this model no longer serves our needs.  It was built for a different world. One where information was scarce, decisions were slow, and data arrived in neat packages that had to be carefully interpreted and passed down the chain.

The constraints we face today are not the same.  Organizations are no longer short of data or the means to communicate it, however, they are short on the confidence needed to make accurate and timely decisions.  Markets are moving so quickly that by the time insight has been gathered or data analyzed, the moment has already passed.

The New Model – Not a Pyramid, Not a Diamond but a Mesh

One emerging view suggests that the next organization model is an evolution of the pyramid into a diamond.  The broad base, traditionally home to a large proportion of junior roles, is shrinking, as AI takes on much of that work. It's an interesting idea and not without controversy but I recently came across an even newer take on what future organizations might look like.

In her new book ‘Refactoring the Firm’ (2026),  Adriana Rocha introduces an entirely new way to think about organizational structures.  She sees future businesses becoming what she calls ‘Intelligence Native Organizations’ (INOs):  less of a fixed shape and more of a flat, fluid mesh.   

Within the mesh, data and information move freely.  AI agents operate in a tightly governed framework – detecting signals, running scenarios, synthesizing inputs and even making decisions whilst we sleep – meaning organizations become truly always-on.

But humans still play a role – an elevated one.  They sit at the edges, they set the vision, the goals and objectives.  When it comes to insight and data, Rocha (2026: 83) puts it simply: “Agents handle the signal. Humans handle the meaning.”  Perhaps even more importantly, humans add the ethical lens – deciding what is good for the business and what actions and decisions reflect the values of the organization.

In this new world, humans aren’t removed but become essential. They “stop becoming the routers of information and become architects of intelligence” (Rocha 2026: 83). Which, if you've been worrying about where you fit in all of this, is rather reassuring.

Agents as Citizens, Not Tools

This is where it gets contentious: Rocha argues that AI agents shouldn't be seen simply as tools but as citizens - embedded in the organization, sitting on the organogram alongside human colleagues, delivering work around the clock (Rocha 2026: 103).

Admittedly, this does sound quite radical, however if these agents are given clear roles and training with hard coded boundaries, inbuilt governance and accessible audit trails then there is no reason why this could not happen.

What Does This Mean for Agencies?

As more companies start to move towards becoming INOs, operating with hybrid intelligence, then others will need to follow or risk being left behind.  An insight or strategy agency with episodic delivery cycles does not fit a client that runs on embedded, always-on intelligence. 

The best agencies won’t run insights projects or ad hoc work but they will operate as an extension of their client’s own infrastructure.  Always absorbing, always analyzing, always looking for the next signal.  Ready to deliver precision intelligence exactly when it is needed.

This requires a different kind of relationship between agency and client – one built on deeper access and longer trust.  It also requires agencies to give something up: the comfort of defined scope, neat deliverables and the invoice that follows.  This is a harder shift than it sounds as it upends the business model entirely, moving from billable hours to pricing based on value.

Not a Distant Concept but an Emerging Reality

Truthfully, I doubt any agency fully operates as an INO today. But five years from now, the gap between those leaning into this and those who aren't will be hard to ignore.

So the question to all insight and strategy agencies: are you a supplier of work, or a partner in decisions?  Only one of those has a long-term future.

References:

Sequoia Capital, 'From Hierarchy to Intelligence', 2025, sequoiacap.com

Refactoring The Firm – Adriana Rocha, Feb 2026