Every agency owner wants to know the same thing: am I running a healthy business? But the benchmarks that get cited most — "$150k revenue per employee," "20% EBITDA margins" — were built on data from before the AI inflection point. We surveyed 340 marketing agencies across the US, UK, and Canada to build a 2026-current picture of what healthy actually looks like now.
The short answer: the goalposts have moved, and agencies that haven't recalibrated are measuring themselves against a benchmark that no longer exists.
The most-cited agency benchmark — $150,000 in revenue per full-time employee — was established when the primary production cost was human time. In 2026, that number has been upended by AI-assisted production, offshore talent integration, and the shift toward retainer-based advisory work.
Our data shows the median revenue per FTE for agencies surveyed is now $187,000 — a 25% jump from where the benchmark stood in 2022. But the distribution tells the more interesting story.
| Agency Tier | Revenue/FTE | AI in Production? | Primary Model |
|---|---|---|---|
| Bottom quartile | < $130k | 18% | Project-based |
| Second quartile | $130k–$175k | 44% | Mixed |
| Third quartile | $175k–$230k | 71% | Retainer-heavy |
| Top quartile | > $230k | 94% | Retainer + advisory |
The correlation is stark: agencies in the top quartile on revenue per employee are nearly universally using AI in their production workflows. This doesn't mean AI alone drives the number — top-quartile agencies have also made deliberate shifts to retainer pricing and advisory services. But the data suggests that AI adoption is a prerequisite, not a differentiator, at that tier.
Agencies crossing $230k revenue per FTE share three traits with 89% consistency: (1) at least 60% of revenue from retainers, (2) AI tools embedded in core deliverable production, and (3) a documented client onboarding process. None of these traits is sufficient on its own.
Vendor surveys about AI adoption tend to skew optimistic — the sample is self-selecting and the methodology often conflates "tried it once" with "using it in production." Our survey asked specifically about AI tools embedded in billable client work, not just internal experimentation.
The results are more nuanced than the "everyone's using AI now" narrative suggests.
That last number — 8% — is worth dwelling on. Despite widespread concern among agency employees about AI displacing jobs, very few agencies have actually reduced headcount because of AI. The more common pattern is that agencies are absorbing more work with the same headcount, or redirecting employee time from production to strategy and client management.
"67% of agencies use AI in production. But only 29% have repriced to account for it — leaving significant margin on the table."
Retention is the single most powerful lever in agency economics. An agency that retains 90% of its clients annually competes in a fundamentally different game than one retaining 70% — even if they have identical new business win rates.
The math is simple: at 70% retention, you need to replace 30% of revenue every year just to stay flat. At 90% retention, you only need to replace 10%, freeing enormous capacity for growth-oriented business development rather than churn replacement.
| Annual Retention Rate | % of Agencies | Median Margin | YoY Growth |
|---|---|---|---|
| < 70% | 21% | 9% | -3% |
| 70–80% | 34% | 13% | 4% |
| 80–90% | 31% | 18% | 11% |
| > 90% | 14% | 24% | 19% |
Only 14% of agencies we surveyed retain more than 90% of clients annually. That group shows a median EBITDA margin of 24% and 19% year-over-year revenue growth — figures that would be considered exceptional in almost any professional services category.
Among agencies reporting over 90% annual client retention, the three most commonly cited practices were: structured quarterly business reviews (78%), documented client health scoring with proactive outreach triggers (61%), and dedicated client success ownership separate from project management (54%).
Agency org structures are in the middle of a significant transition as AI reshapes production roles. Here's what current team structures look like across agency size:
The median team size is 8 people, with a ratio of roughly 60% producers (writers, designers, strategists doing billable work) to 40% account/operations. AI adoption is highest among the owner/leadership layer, with usage often concentrated in content production and reporting automation. Average utilization rate: 71%.
This is where team structure diverges most sharply based on model. Project-based agencies in this tier average 22 FTEs and show utilization of 68%. Retainer-focused agencies at the same revenue level average 16 FTEs — 38% more revenue per person. The gap is primarily explained by lower scoping overhead and more predictable work rhythms on retainers.
These agencies increasingly look like two distinct categories: those that have rebuilt their production layer around AI and offshore talent, achieving margins of 20–28%, and those that haven't, where margins are compressing toward 10–14% as competitive pricing pressure grows.
Industry standard for healthy utilization is 65–75% of billable capacity. But top-quartile agencies measure differently — they track realized value per hour rather than raw utilization, and many run intentionally lower utilization (55–65%) to protect strategic thinking time and reduce attrition. High utilization and high margin are not the same goal.
New business development remains the highest-anxiety function in most agencies, and the benchmarks here are some of the most misunderstood in the industry.
The referral number is the most telling. At the agencies growing fastest in our dataset, nearly three-quarters of new revenue came through referrals or inbound from existing clients and their networks. This isn't a "sales problem" — it's a client satisfaction and visibility problem. Agencies that solve for retention and build content-driven visibility don't need large outbound sales motions.
Across every metric we cut, the gap between top-quartile and median agencies has widened since 2022. The agencies at the top share a recognizable profile:
None of these are new observations. But the gap between agencies that have executed on this list and those that haven't has grown significantly in the last three years. AI has been an accelerant: it rewards agencies already running efficient, retainer-heavy models, and compounds the disadvantage for those that haven't made the structural shift.
A top-quartile agency at $230k revenue per FTE, 91% retention, and 24% EBITDA margins isn't just more profitable today — it's compounding faster. Every retained client reduces next year's new business burden. Every AI efficiency gain funds capacity for more strategic work. The structural advantages are self-reinforcing.
A few notes on applying this data to your own agency. Benchmarks are diagnostic, not prescriptive. A 70% retention rate at a project-based agency doing event work is structurally different from 70% retention at a content retainer agency — the client relationship dynamics aren't comparable. Use these numbers to identify where you're meaningfully above or below cohort peers, then investigate the why before drawing conclusions.
Also worth noting: averages obscure bimodal distributions. In several of these categories — particularly AI adoption and margin — we're seeing a "barbell" pattern, with a cluster of high performers and a cluster of laggards, and fewer agencies in the middle than two years ago. If you're in the middle, that's actually the most precarious position: not yet differentiated as a high-performer, but with cost structures that make it difficult to compete on price with lower-overhead operations.
The direction of the data is clear. The agencies investing in retainer models, AI-assisted production, and active retention management are pulling away from the pack. The window to make these structural changes while the market is still relatively stable — before the next round of client budget pressure — is finite.
The complete report includes 60+ data points across revenue, profitability, team structure, AI adoption, client retention, and new business — with breakdowns by agency size, specialty, and geography.
Download Full Report — $97