The Business Model Crisis: Why the Agency Model Is Breaking

Part 4 of 6: The Structural Crisis in Public Relations

Why the PR Agency Model Is Breaking, and the Problem Is Structural

Here is the uncomfortable math that most agency leaders know but rarely say out loud: when your largest client internalizes communications, you don't just lose revenue. You lose the economic structure that made your entire model viable.

The structural forces described in the first three parts of this series (fractured attention, contextual trust, measurement pressure) don't stay abstract for long. They show up in the P&L. They show up in the client roster. And they are forcing the PR agency model into a set of contradictions it was never designed to resolve.

Why the whale clients are leaving

Large organizations are not abandoning public relations. They are internalizing it.

As communications environments become continuous rather than episodic, major clients increasingly need PR to function as an embedded capability: always present, deeply informed, tightly aligned with business priorities. Managing real-time narratives across social platforms, monitoring emerging risks, coordinating messaging across teams, intervening before issues escalate. These require constant context and proximity. They are difficult to meet through an external model designed around periodic engagement.

At the same time, the declining marginal impact of traditional media placement has weakened one of the agency model's historical advantages. In a fragmented attention economy, audiences reside in segmented cultural and digital spaces beyond the reach of any single outlet. For large clients with the resources to build, the logic increasingly favors internal teams that operate continuously rather than outsourcing reactivity.

This trend isn't speculative. Agencies are absorbing up to five times more client load to offset the loss of their anchor accounts. The top 250 PR firms generate $18 billion in combined fee income, and the margin pressure is structural, not cyclical. When whale clients leave, agencies don't just lose revenue. They lose the leverage that subsidized everything else.

The headcount trap

Historically, whale clients gave agencies leverage. A small number of large retainers supported senior judgment, absorbed inefficiencies, and stabilized margins. When those clients internalize, agencies are forced to replace concentrated revenue with volume.

On paper, this looks like diversification. In practice, it's a trap.

PR capacity scales with people. Each additional client brings its own coordination costs, reporting expectations, approval cycles, and service rhythms. Replacing one large client often requires onboarding several smaller ones, which requires hiring more staff to maintain service levels. Costs rise faster than revenue. Margins compress. Operational complexity increases.

The result is a fragile equilibrium. Agencies become busier but not healthier. Revenue may be preserved in the short term, but baseline costs (salaries, overhead, coordination load) rise permanently. In downturns, this exposure becomes acute: salaries must still be paid even as client churn increases and replacement revenue grows uncertain.

What appears to be diversification is, in fact, a loss of leverage.

Why more people no longer mean more value

The headcount trap is compounded by a deeper problem: additional labor does not translate proportionally into client value.

A significant share of agency effort is consumed by activities that are structurally necessary under the traditional model but do not directly improve outcomes: pipeline management, internal coordination, reporting rituals, status updates, relationship maintenance. These activities are often justified as "relationship building," an orthodoxy that has long defined agency culture. The assumption is that proximity, cadence, and partnership experience are themselves sources of value.

Clients increasingly disagree.

What agencies perceive as relationship investment is often experienced by clients as overhead. Trust and alignment still matter, but they are no longer sufficient to justify cost on their own. Clients pay for results: influence, engagement, momentum, risk mitigation. Not for process visibility. The tolerance for effort that can't be clearly linked to outcomes is shrinking fast. Only 13% of in-house teams that work with agencies say the partnership is going well, according to Superside's 2025 survey. Fifty-one percent have lost faith in agencies entirely.

This creates a structural contradiction. Agencies respond to revenue pressure by increasing headcount, but much of that capacity is absorbed by maintaining the machinery of the agency-client relationship rather than advancing the client's objectives. From the client's perspective, output doesn't improve in proportion to effort. From the agency's perspective, costs rise faster than value delivered.

Under these conditions, scaling through headcount alone is not just inefficient; it actively undermines both margin and credibility. What once felt like "white-glove service" increasingly registers as misallocated effort in a results-driven environment.

The efficiency ceiling

Some agencies have tried to address these pressures through process optimization: better project management, tighter SOPs, more structured workflows. These help at the margins. They do not solve the structural problem.

The core issue is that human effort is being consumed by coordination work that doesn't create proportional value, in an environment that demands more judgment, more speed, and more contextual sophistication than ever. You can optimize a workflow, but you can't optimize your way out of a model where revenue is fundamentally tied to the number of people on staff.

The math is straightforward: if every additional client requires additional people, and each person brings fixed costs regardless of utilization, then the agency model has a built-in ceiling on both margin and scalability. Process improvements can push that ceiling up slightly. They cannot remove it.

Meanwhile, the demands described in Parts 1 through 3 keep intensifying. More channels to monitor. More trust contexts to navigate. More pressure to prove ROI. More speed. More sophistication. All of it requiring human judgment, which is the one thing being consumed by the coordination tax of the model itself.

Professional services like PR derive their value from judgment-heavy work: interpretation, prioritization, timing, framing, and restraint. Yet the reality of running agencies means that a disproportionate share of human effort is absorbed by work that neither clients value nor professionals enjoy. Administrative overhead crowds out the very expertise clients are paying for.

What this means

Public relations is not disappearing. The discipline is not irrelevant. The structural forces described throughout this series make skilled communications more necessary than at any point in the last thirty years.

But the model built to deliver that work (the model built on episodic engagement, manual process, and human effort allocated to administration as much as to strategy) is reaching its limit. Agencies face an unsustainable tension: deliver internal-level awareness, speed, and accountability while charging clients for time spent navigating bureaucracy. Clients are no longer willing to fund that inefficiency. They expect people to be working on problems that actually move the needle.

The firms that adapt will not win by hiring more people. That lever is exhausted. Something structural has to change, not in the capability of the people, but in how their effort is allocated. The question is no longer whether the agency model needs to evolve. It's what the next model looks like, and what force is powerful enough to redraw the boundary between work that requires human judgment and work that never should have required it in the first place.

Jessen Gibbs is CEO and Co-Founder of Shadow, an AI-native PR agency built by embedding inside the world's best communications firms.