Redesign the Floor: What Changes When a Communications Team Is Rebuilt Around AI

What Changes When a Communications Team Is Rebuilt Around AI

When factories first replaced steam engines with electric motors, productivity barely moved. Economic historian Paul David, who studied the transition, found that owners kept their old multistory layouts, built around a central power shaft. The real gains took decades to arrive. They came only when someone redesigned the floor around what electricity made possible.

Communications is in its electric-motor moment. In BCG's 2026 survey of 300 chief marketing officers, 96 percent say AI is driving end-to-end transformation of their function, while only about a third have moved beyond the basics. The tools are in the building. The work has not moved.

I have watched that gap from both sides. At Gutenberg Communications, the agency I founded, more than 100 people across seven countries rebuilt how they work. At CambrianEdge.ai, we built the environment that made it possible, then paid close attention to what it took. The answer was not a better tool. It was a different way of working, and getting there was a leadership challenge long before it was a technology one.

The enemy is disconnected work

Gutenberg's early experience with AI looked like that of most organizations: capable people, many models and teams in different time zones, each using a different tool in a different way. Every new application added another interface and another place where context had to be recreated. AI was supposed to simplify the work. Instead, it was adding another layer.

The problem was never the technology. It was disconnected work. So we did not ask people to bolt AI onto existing processes. We rebuilt the workflow so that research, strategy, creation, review and distribution moved together, with context traveling alongside the work. According to Gutenberg's reported results, every one of its 100-plus employees now works this way. Execution that once took weeks now moves in days, roughly five times faster, and about nine in 10 first drafts are approved.

What had to change

Choosing the technology took weeks. Changing habits took longer, and four shifts mattered most.

Resistance arrived as waiting, not refusal. Nobody declined to use AI. Capable people waited for proof, for permission, for a quieter quarter. Mandates changed little. Progress came when leaders stopped announcing tools and began redesigning one painful task at a time, alongside the person who owned it. People adopt a new way of working when it solves a problem they already resent.

AI literacy had to become AI fluency. Knowing how to use a tool is literacy. Knowing when to trust its output, when to overrule it and how to redesign a workflow around it is fluency. It is the capability clients will pay for. Every Gutenberg employee completed a structured AI literacy program. AI fluency, though, is built through real work, not in a classroom.

Leaders had to review differently. Approval layers and weekly review cycles, once the machinery of governance, became friction. The leader's job moved from inspecting every output to setting the standard before the work begins: the brief, the voice and the lines that cannot be crossed. We also adopted a discipline I recommend to any team. A long AI-generated document opens with an executive summary that leadership can accept or reject. If the summary does not hold, the other 30 pages do not get read.

Judgment had to stay visibly human. AI can produce material faster than any team can use it, which makes editorial judgment the scarce resource. The risks worth managing are the ones a leader can govern directly: accuracy, consistency of voice and clarity about who approved what. AI accelerates the work. People direct it, review it and approve it, and everything that leaves the team carries a human decision.

Structure follows the work

Gutenberg now operates in pods that bring planning, content, video, search and answer-engine optimization, media, social and performance analytics into one linked workflow. The harder change was conceptual. A reporting line is a military idea: authority flows down and everyone waits for the order. A pod works more like an octopus. Every arm can think for itself, and one coordinating intelligence keeps the movement coherent. The orchestrator sets direction and alignment. Nobody waits for permission to use their judgment.

Roles moved in a consistent direction, from passing work along a chain to owning an outcome. Specialists who once handed a brief from department to department now sit around the client together, sharing the same context. The working rule is simple: AI drafts, people decide. Research, first drafts, targeting and reporting are accelerated. Relationships, narrative and the final call stay with people.

From faster work to new products

A rebuilt workflow can be packaged. Once research, creation, review and distribution move together, the output is repeatable enough to price and sell as a product rather than a stack of hours, helping protect margins as the work speeds up. At Gutenberg, video production was the first service rebuilt this way. AI visibility is the second, and it is where communicators have the stronger hand.

McKinsey found that a brand's own sites account for only 5 to 10 percent of the sources AI search draws on, while Muck Rack's analysis attributes about 84 percent of citations to earned media. The sources AI draws upon are the ones PR has always worked to earn. For a communications team, that turns an existing skill into a new line of revenue.

What senior communicators should prioritize now

First, find the workflow that costs your team the most and rebuild it end to end before buying anything else. A tool nobody changes their week for is a subscription, not a strategy.

Second, protect your freedom to change models. No single model will stay best at every task, so build around the workflow, not the vendor. Your workflow should outlive your AI model.

Third, invest in fluency with a standard behind it, not a one-off workshop. Clients have read the same headlines about AI lowering the cost of production, and they will ask where the savings went. Teams that can show how they use AI, and how they govern it, will have a better answer than a defense of the hours. They will scale client work, not coordination overhead.

From "How I AI" to "How We AI"

AI adoption moves in acts. The first is personal: how I AI, one professional and one prompt at a time. The second is organizational: how we AI, where habits, structure and trust decide the outcome. Most communications teams are standing at the end of the first act, wondering why the second feels so hard. It is hard because it is a leadership act.

The motors are installed. Now redesign the floor.

Harjiv Singh

AI Overview

Harjiv Singh is a New York-based entrepreneur and marketing executive best known as the founder of the global communications agency Gutenberg and the CEO and co-founder of the AI-native marketing platform CambrianEdge.ai.

Next
Next

When the CEO Becomes the Crisis, Can the Brand Stand on Its Own?