The Conversational Workplace

Here is the honest version of how most agencies actually work. Strategy writes the brief. Creative receives the brief. Production executes the brief. Accounts manage what falls out of that sequence. Everyone nods along at the all-hands about how collaborative they are. Then they go back to their desks, inside their functions, and do the same thing again next week.

This is not a cynical portrait. It is a structural one. And the structure is old. Older than Slack. Older than open-plan offices. It is essentially an assembly-line model wearing creative-industry clothing, and it survives because it is efficient, legible, and deeply comfortable for everyone who has organized their professional identity around owning a specific piece of the process.

The problem is not that it fails to produce work. It produces plenty of work. The problem is what it costs the work — quietly, repeatedly, in ways that are genuinely difficult to trace back to the source.

Conway's Law, established by Melvin Conway in 1967, states that organizations design systems reflecting their communication structures. Although initially about software architecture, it applies broadly: the design mirrors internal conversations. If conversations run in parallel, work does too. If strategy and creative seldom communicate until the project is done, the campaign feels disconnected. Clients notice this but often lack the term for it.

A 2022 McKinsey survey found that most organizations rated their internal collaboration positively at the individual level while simultaneously identifying siloed functions as one of their leading structural barriers to performance. Which means the people are fine. The system is the problem. And in an industry where the most important variable is the quality of the thinking before a single word is written or a pixel is moved, a system that keeps disciplines in their own lanes is not a neutral operational choice. It is a creative tax paid on every project.

So what would the alternative actually look like? Not in values-statement language. In operational terms.

A conversational workplace is not an organization that talks more. Most organizations already have too many meetings. It is an organization where ideas move across functions early enough to actually change things, where the strategist's hypothesis is live material, available to the creative team before it has hardened into received wisdom, and where the designer's instinct about a direction can push back on the strategy before the team has committed to it emotionally. Where the account lead's knowledge of the client's actual anxieties — the ones that live below the brief — enters the strategic conversation rather than being filtered into a feedback loop at the end.

MIT's Sloan School has a phrase for this kind of organizational capacity: integrative intelligence. The ability to combine knowledge across domains rather than simply deepen expertise within them. The concept sounds abstract until you watch a campaign come apart because the media planner and the copywriter never spoke until the work was already in production. The concept becomes very concrete very quickly.

The jazz ensemble metaphor gets used a lot in this conversation, probably because it earns its place. An orchestra plays from fixed sheet music; every section waits for its cue, executes with precision, and the result is beautiful but predetermined. A jazz ensemble is a different organism entirely. Everyone is listening. Everyone is responding. A phrase from the trumpet changes what the piano does next. The result is not chaos. It is structured improvisation, and it produces something that no single player could have produced alone, and that could not have been written in advance.

The agency equivalent of that difference is not about personality type or office layout. It is about when the conversation happens and who is invited into it.

There is a distinction worth pressing on here because it is where a lot of agencies fool themselves. The distinction between coordination and collaboration. They are not synonyms, and treating them as such is one of the more elegant organizational self-deceptions available.

Coordination is the management of workflow. It is task assignment, timeline tracking, handoff documentation, and the infrastructure that keeps a complex process from becoming a disaster. Coordination is genuinely necessary and often undervalued. But it is essentially administrative. It governs the movement of work through a system. It does not, on its own, improve the thinking inside that work.

Collaboration is something else. It is the shared development of an idea across different kinds of intelligence. It requires that perspectives are brought into genuine contact before decisions calcify — that the strategist is changed by what the creative director said, that the brief is sharper because the writer pushed back on a premise, that the final work carries the fingerprints of multiple disciplines because those disciplines actually shaped it. Not just executed their respective portions of it.

The diagnostic question is simple enough. When the work is done and is genuinely exceptional, what made it that way? If the honest answer is always "because of the brief" or "because of the creative concept," the organization probably has good coordination and occasional collaboration. If the honest answer is more complicated — if it involves a conversation that happened before anyone knew what the work would be, a tension that got surfaced early enough to improve the strategy rather than just survive the review — that is a conversational workplace operating as designed.

Amy Edmondson at Harvard Business School spent years studying what she calls psychological safety: the shared belief within a team that taking an interpersonal risk will not result in punishment or humiliation. Her research, originally conducted in hospitals where the stakes of intellectual silence are measurable in patient outcomes, consistently shows that psychological safety is the strongest predictor of team learning behaviour. And team learning behaviour is what produces the kind of work that surprises people.

Google ran a multi-year internal research project — Project Aristotle — to identify what distinguished its highest-performing teams from its merely competent ones. The hypothesis going in was that the best teams would be composed of the most talented individuals. The finding coming out were that individual talent was less predictive than the team's psychological safety. What mattered most was whether people felt safe to bring incomplete ideas into the room and have them developed rather than dispatched.

This is the cultural foundation without which everything else in this article is aspirational noise. You cannot build a conversational workplace on a culture of intellectual performance anxiety. People who are afraid to be wrong in public will wait until they are certain before they speak, and by then the window for genuine influence on the direction of the work has usually closed. The organizations that produce consistently interesting work are almost always ones where senior people are visibly curious, visibly uncertain, and visibly interested in what people more junior than them are thinking. Not as an affectation of approachability. As an actual mechanism for accessing better thinking.

The Pixar model adds something useful here. Their internal creative process uses a technique called "plussing" — the principle that you can only critique an idea if you also add something constructive to it. Not because the culture is precious about criticism. Because forward-moving critique keeps the conversation alive, and conversations that stay alive produce better outcomes than ones that get shut down by ego or hierarchy. The "hero creative" archetype — the single gatekeeper through whom all good thinking must pass — is the specific failure mode this principle guards against. One person's taste, however excellent, cannot substitute for a culture of shared intelligence.

Francesca Gino's research at Harvard on curiosity in the workplace found that organizations with higher levels of expressed curiosity showed lower conflict rates, better information sharing, and measurably higher creative output per employee. The mechanism matters: curious people are more likely to ask the questions that surface knowledge others are holding but not offering. They are more likely to find the assumption in the strategy that nobody has tested yet, the client insight that has been living in the account lead's head for three years without ever making it into a brief.

Cognitive science frames this through the concept of conceptual blending. The most generative ideas tend not to emerge from deep immersion within a single domain. They emerge from the encounter between different conceptual frameworks — from the moment a strategist's model meets a designer's intuition, and neither is quite the same afterward. This is not an argument against expertise. Deep expertise is irreplaceable. But expertise without exposure produces teams that are technically excellent and intellectually isolated, and isolated solve bounded problems well while missing the more interesting, more difficult, more valuable questions entirely.

The Harvard Business Review has documented this in agency environments specifically. Teams with regular cross-disciplinary exchange produce work that clients rate as more original and more commercially coherent than teams with higher individual talent scores but lower rates of genuine internal dialogue. The finding is counterintuitive if you believe that the best work comes from the best individuals. It is completely predictable if you believe that the best work comes from the best conversations.

What does it take, in practice, to build the kind of organization where this happens reliably rather than accidentally?

Earlier involvement is probably the highest-leverage change available. Moving the moment of cross-disciplinary contact upstream — into the briefing stage rather than the review stage — changes the nature of every conversation that follows. When the creative director is in the room as the strategy is being built, not to approve it but to stress-test whether it will resolve in execution, the strategy tends to become more useful. When the writer is present as the platform is framed, the language choices in the brief become more precise. These are not marginal improvements. They compound.

The brief itself is a significant variable. A brief written as a closed document — a set of instructions implying that the thinking is complete — naturally suppresses conversation. A brief written as a shared problem, one that names its own uncertainties and invites interrogation of its framing, creates a different relationship between everyone involved. The brief becomes the start of the conversation rather than the end of the thinking.

The structure of creative critique is another lever that most agencies underuse. When a review functions as an approval process — work is presented, judgment is rendered, work is revised — the organization learns to manage critique rather than think through it. When a review functions as a collective thinking exercise, the question shifts from "does this work?" to "what is the thinking behind what is strongest here, and what does that open?" That kind of conversation is harder to run and slower. It is also the kind that makes everyone in the room better at their job.

Leadership modelling determines whether any of this is sustainable. When senior people demonstrate intellectual curiosity visibly — when they ask genuine questions in rooms where they could simply provide answers, when they change their position because of what someone more junior said, when they are seen to not know things and be interested in that — they give the entire organization permission to operate the same way. Culture in most organizations flows downward with a reliability that no communication strategy can match.

The most compelling agencies are not merely efficient or pleasant or well-staffed. They are intellectually alive. The work they produce reflects not just individual talent but a quality of collective thinking that is genuinely difficult to replicate from the outside, because it lives in the habits and conversations and cultural permissions that exist inside the organization.

Clients who have experienced this kind of agency know the difference. The work lands differently. It has a coherence that goes deeper than visual consistency or strategic alignment — a quality that comes from the fact that the thinking behind it was genuinely shared, that multiple kinds of intelligence helped shape what the problem actually was before anyone tried to solve it.

The organizations that build this do not do so by hiring collaborative people and hoping for the best. They build systems that reward the right conversations, structure the right moments of contact, and lead in ways that make intellectual generosity the obvious professional choice. They treat conversation not as a cultural amenity but as the mechanism through which better work is reliably made.

The highest compliment a team can pay its workplace is probably not that it is easy to work there. It is the people in it who make your ideas better. That is a different kind of culture entirely, and it is worth building deliberately.

‍