AI and the Future of Agencies: Reinvention or Smokescreen?
AI is forcing a reset across marketing. Some leaders call it the next creative growth engine. Others say it's a handy cover for cost-cutting. Both can be true at the same time.
Here's the signal from dozens of senior marketers: leaner teams, faster output, tighter margins, and a premium on work that feels undeniably human. The question isn't "Will jobs go?" It's "Who adapts fast enough to stay valuable?"
What's actually changing inside teams
- Smaller, senior-heavy teams are shipping at a pace that once needed full departments. AI lets one person do the work of three.
- Clients are pickier. They expect time-to-value on AI work and a clear "why" behind your approach, not just cheaper deliverables.
- Work is shifting from one-off executions to systems: repeatable creative, data feedback loops, and measurement built in from the start.
As one CTO put it, growth with less budget would've sounded impossible a few years ago. Now it's table stakes-if your team knows how to use the tools.
Where cuts will land (and where they won't)
Job losses are real in pockets where tasks are easy to automate. Junior roles are exposed first as templates, agents, and automation take on repetitive work.
Large networks feel the squeeze the most: high cost bases, scrutiny on media margins, and merger fallout. Mid-size independents are holding steady by staying lean and adding a few new specialist roles.
Net-net: total industry headcount likely dips, while smaller and mid-sized shops win share. In-housing marches on as clients build their own AI muscle and buy fewer hours.
The two agency paths (and the gap between them)
- The cost-cutter: treat AI as headcount reduction. Margins get a short-term bump, capability erodes, and clients wonder why they still need you once they have the same tools.
- The capability builder: use AI to free talent for higher-value work-original strategy, concept development, brand design, and "proof of life" content that no model can fake.
Christophe Jammet calls it "proof of life marketing." In a feed stuffed with AI slop, the work that feels unmistakably human earns the click-and the budget.
Beware the efficiency theater
Many "AI-driven efficiency" stories are spin to mask cuts. Yes, automation saves time. No, it doesn't magically double profit. In practice, AI often helps you hold margins while formats multiply and production cycles speed up.
That's not a failure-that's the new baseline.
What clients will pay for next
- Proprietary advantage: unique data, creative frameworks, or market knowledge that doesn't come out of the box.
- System outcomes: not just an ad, but the flywheel-audience model, creative engine, and measurement loop that improves each sprint.
- Proof of ROI on creativity: clear tests, incremental lift, and time-to-value that finance can underwrite.
As one founder warned, clients will soon have access to the same AI and media buying power. If your edge isn't defensible, you're the middle layer.
Roles that are rising
- AI producer: pairs creative taste with model know-how; turns prompts, references, and iterations into usable outputs.
- QA and brand safety lead: sets standards, checks for bias, accuracy, and compliance.
- Data partnerships lead: builds the data "moat" and integrates it into creative and media systems.
- Client educator: translates AI capability into business cases the CMO can defend.
April Quinn's take: disruption first, then opportunity-if you re-skill and set a higher bar for craft.
Don't gut your junior pipeline
There's a real risk of skipping a generation. Cutting all junior roles feels efficient-until you need new voices, fresh references, and future leaders.
One agency leader keeps 25% junior talent by design. Pair them with seniors. Build apprenticeship loops. Your culture and your future self will thank you.
Pricing and in-housing: accept the pressure
Prices will fall as adoption spreads. Clients will take more inside. Agencies with deep expertise will still thrive by pairing AI with high-value services-brand strategy, original platforms, and creative that creates a social signal AI can't mimic.
Think "more, better, faster"-not just cheaper. Kate Ross framed it well: jobs are bundles of tasks. Automation unbundles and rebundles the work. Most modern roles evolve rather than vanish.
Practical plan: 90 days to real AI impact
- Audit: map tasks by frequency, time spent, and risk. Automate low-risk, high-volume work first.
- Policy: write a plain-language AI use policy (privacy, attribution, approval). Make it easy to follow.
- Skills: identify gaps. Train producers, creatives, and account leads on the specific tools tied to your workflows.
- Proof of life: ship one human-led, AI-assisted flagship piece per quarter (film, live concept, long-form story) that signals brand taste.
- Measurement: agree on a two-metric ROI model per project (e.g., cost per asset + creative lift). Report weekly.
- Data edge: secure at least one data partnership that your competitors don't have. Build it into your creative system, not just media.
- Client pilots: run two AI pilots with guaranteed time-to-value and a clear "stop/scale" rule.
Metrics that matter now
- Time to first draft and time to final (by asset type).
- Cost per concept and hit rate (ideas that make it into market).
- Throughput per FTE without quality loss (measured by engagement or brand-lift proxy).
- Percentage of human-only assets in each campaign (your "proof of life" ratio).
- Attribution confidence for AI-assisted work (how sure are you that it moved the metric?).
What the data says (and what to do with it)
Some analysts predict AI-related workforce reductions sooner than expected. Others point to job growth as roles rebundle and new specialties emerge. The truth is likely uneven: cuts in some places, growth in others.
- Global outlook: The World Economic Forum anticipates shifts in job growth by 2030 as tasks rebundle. Read their latest report.
- Enterprise impact: Research firms expect automation to reshape headcount and skills mix over the next few years. See analyst coverage.
Agency playbook by size
- Large networks: simplify your offer, kill low-margin complexity, and put proprietary data and frameworks at the core. Prove ROI on creativity, not just media.
- Mid-size independents: double down on senior pods and system-based scopes. Build 1-2 durable moats (data, format expertise, category depth).
- Boutiques: become the specialist clients can't replicate in-house. Lead with taste, distinctive IP, and speed.
Where this is heading
Automation will trim certain tasks. Clients will expect more, faster. Agencies that thrive will couple AI with judgment, taste, and a measurable system for creating value.
As multiple leaders echoed: direction still comes from human taste, instinct, and ambition. The next era won't be measured by billable hours-it will be measured by outcomes people actually feel.
Level up your team's AI capability
If you're building skills across creative, media, and strategy, start with focused training your team can apply this week:
- AI courses by job function for role-specific upskilling.
- AI certification for marketing specialists to standardize best practices across the team.
Bottom line
AI is not a magic margin machine. It's a leverage tool. Used well, it frees your best people to focus on high-value thinking and work that proves there are real humans behind the brand.
Call it what you want-just make sure your output makes the case.
Your membership also unlocks:
