We’re all dealing with the same tension. Let’s call it out: we’re being asked to move mountains with fewer people, bigger goals, and a growing pile of AI tools that supposedly solve everything. It’s exhausting — and nowhere is it more obvious than in our GTM teams.
The good news? Engineering has been here before. Since the early days, they’ve had to move mountains with fewer people, tighter budgets, and massive expectations.
(audible gasp) “Kyle, are you suggesting we talk to our engineering team?”
Yep. Because marketing — especially product marketing — is next. We all love talking about revenue, efficiency, and sales velocity, but PMM is usually the first team to feel the flattening. But, their work drives quota attainment and sits at the center of whether GTM wins.
They’re essential.
Let’s look back at our friends in engineering, starting with Adam Ferrari. I read his Substack often (we worked together at Jellyfish), and he recently wrote The Great Flattening, in which he argues that companies are stripping out layers due to two forces: efficiency pressure and AI.
Said another way: “How much middle do you really need?”
That lens maps cleanly to PMM. For years, product marketing looked like 2010s engineering — bigger teams, heavy specialization, and lots of translation between product, sales, and customers. It made sense when product cycles were slow and information was hard to gather.
Now the model is changing.
Efficiency pressures: As engineering shifted from “growth at all costs” to “efficiency at all costs,” PMM teams are slimming down. Instead of a PMM per product, you now see lean pods covering messaging, enablement, and GTM strategy across portfolios.
The AI effect: AI eats a lot of management toil. In product marketing, it drafts messaging frameworks, summarizes calls, assembles competitive briefs, and surfaces early win–loss signals. One strong PMM can now cover the surface area of three.
The role shift: The best engineering managers go beyond updating Jira — they unblock people and systems. Same for PMMs. Their highest value isn’t a battle card or datasheet; it’s shaping the category story, influencing the roadmap, and coaching revenue teams. AI handles the busywork. The human side makes PMM irreplaceable.
Adam ends with a warning: flatten too far and things break. A 10:1 engineer-to-manager ratio looks great on paper until quality and coaching collapse. Product marketing is no different. Cut too deep and you lose the tissue that connects product, sales, and the market. Engineers build features nobody sells. Sellers tell stories nobody cares about.
So how do we flatten with intention?
1. Align PMM to outcomes. We still organize by product line, but each PMM owns outcomes — mainly revenue. They’re deeply aligned with design, ops, sales, and other teams to speed decisions and reduce friction.
2. Make experimentation part of the job. We run two-week GTM sprints with a clear hypothesis, metric, and decision rule: keep, kill, or scale. Every marketer runs two AI projects per quarter, and we review wins and misses together.
3. Use AI to raise the floor. We’ve standardized a simple stack for first drafts, call summaries, clustering feedback, and fast variants. That frees the team to focus on narrative, influence, and system design — the human stuff that actually matters.
The key to all of this is AI and experimentation. So how are we doing it?
Dozens of AI Projects in Eight Weeks
At Docebo, we ran more than a dozen AI projects in eight weeks. Only one was a clear winner. Some were wins, some flops, all learning experiences.
A few examples:
-
Audience insights from calls (UnifyApps)
-
Personalized account pages (Tofu)
-
AI-powered website copy (Webflow)
-
Revbot for on-the-go insights
-
AI reporting through Dreamdata and Hockeystack
-
AI-driven homepage copy and a site bot trained on our data
Only one was a homerun — our Audience Insights tool built through UnifyApps. It analyzed 4,000+ calls, surfaced objection trees and trigger events, and fed directly into messaging, enablement, and product feedback. Paired with SEO and GEO tools, it lets us focus on speed instead of waiting for perfect ROI.
Don’t chase cuts; chase faster learning, clear ownership, and a story that shows up everywhere.
What didn’t work (yet):
-
AI reporting (data still too messy)
-
Website bot (needs more value)
-
Copy experiments (numbers flat)
The real lesson from the great flattening: the win wasn’t a tool — it was velocity. Learning moved faster across product, sales, and marketing, which is what flattening is meant to buy you.
And what actually works is usually boring. Standardize a small AI stack for call summaries, drafting, and analysis. Write down your prompts. Share patterns. Make the basics run on rails so your team can spend its time on story and decisions.
Bottom Line: Don’t Chase Headcount Cuts
Adam Ferrari is right: the middle is shrinking due to efficiency and AI. Product marketing and the rest of GTM are on the same track. But don’t chase cuts; chase faster learning, clear ownership, and a story that shows up everywhere.
Tools won’t set you apart. Your system will. And your system is only as strong as the people running it. Or as my boss Alessio Artuffo says: “In a world where everyone has AI tools, your only competitive advantage is your people.”
