Zeta Global president Steven Gerber and chief growth officer Ed See explain why, as AI makes creation easier, identifying what’s worth creating matters most.
Marketing has an abundance problem.
AI can now produce in minutes what teams once fought quarters to create: more copy, more segments, more variants, more analysis, more tests. Production is up everywhere. Proof is not.
When almost anything can be produced, the harder question is what is worth producing.
Every era creates a new center of value creation and a defining talent archetype. The software era made engineering central and gave us the 10x Engineer. The AI era is doing the same thing to marketing. Its archetype is the 10x Marketer: not the marketer who produces 10 times more output but the marketer who helps the organization make 10 times better decisions—who to reach, what to say, where to spend, when to act, what to stop, and how to prove what worked—and builds the systems that make those decisions repeatable.
AI does not simply change how marketing works. It changes what marketing is. Production functions create outputs. Intelligence functions improve decisions that drive growth.
Marketing is becoming an enterprise intelligence function, where customer understanding shapes business decisions. The 10x Marketer builds the judgment and systems that make that capability repeatable. The challenge is developing more of them.
From production function to intelligence function
Each era has expanded what marketing can scale. Analog media delivered mass reach. Digital technology enabled targeted, personalized communications across an exploding number of channels. AI makes production abundant, and in doing so, shifts marketing’s advantage from producing more to deciding better.
The central question shifts from “What can we make?” to “What deserves attention, investment, and customer exposure?”
The central question shifts from “What can we make?”
That is not a workflow change. It is a change in what the function is for. The scarce resource is judgment, and few functions sit where marketing does, at the intersection of customer understanding, experimentation, capital allocation, and proof.
The operating model is still catching up with the ambition. Boston Consulting Group’s 2026 survey of 300 global chief marketing officers found that 96% say AI is transforming marketing end to end, yet only about one-third have moved to agent-led workflows. Even that measures how work gets done, not whether better decisions follow. The opportunity is to build an intelligence function whose success is measured by the business value its decisions create.
Consider a specialty retailer we worked with. Several acquisition campaigns were posting strong return on ad spend, so the natural instinct was to scale them. Instead, an analyst connected campaign performance with returns and cost-to-serve, testing whether those campaigns were creating value for the business. That changed the picture. Some of the campaigns that appeared to be winners were acquiring customers who became unprofitable once those downstream costs were factored in. Rather than scale what looked successful on the surface, the team shifted investment toward lifecycle, loyalty, and cross-sell programs with stronger marginal returns.
The breakthrough was not more output. It was a better investment decision.
The apprenticeship paradox
AI is quietly eroding your marketing leadership succession plan. That is the tension inside the transition: AI rewards judgment, but it is dismantling the system that produced judgme
The execution layer of marketing is collapsing—producing, translating, routing, reviewing, reporting. Fewer handoffs, faster cycles, less repetitive work. But execution was also the apprenticeship. Junior marketers learned taste by writing copy, seeing it rejected, and eventually understanding why. They learned judgment in reviews where brand, legal, finance, and customer reality collided. They learned proof by watching campaigns that looked good in isolation fail in the business.
If AI absorbs too many of those reps, companies may move faster today while weakening the judgment they will need tomorrow. An intelligence function that cannot develop intelligent marketers is a function in liquidation.
The 10x Marketer develops three capabilities:
- Taste: Knowing why one of a thousand options is right and saying no to the rest
- Judgment: Framing the right problem before asking AI for an answer and turning constraints into guardrails rather than bottlenecks
- Proof: Establishing whether a decision created incremental business value and using that evidence to guide the next investment
Speed without taste is mediocrity at scale. Speed without judgment is risk at scale. Speed without proof is activity at scale.
The organizational problem is that output is easier to measure than judgment. Campaigns launched, assets produced, and variants tested are visible. A decision that prevents wasted investment is harder to recognize—a great “no” rarely gets celebrated. When promotions favor visible activity over demonstrated business value, people learn to produce more rather than decide better. The reward system keeps training people for the job as it was.
Rebuilding the ladder as a loop
As execution provides fewer learning opportunities, companies must create new ones deliberately. Start with a decision log: Record consequential choices, expected outcomes, and a one-line why. Have junior marketers form a recommendation before senior review, so they practice judgment rather than simply observe it. Then revisit results together: Which assumptions held, which failed, and what should change next time? The log becomes a learning loop when outcomes are used to challenge the reasoning behind decisions, including those made by senior leaders.
The goal is not autonomous marketing. The goal is precision at scale. Marketing has always aspired to understand every customer, every moment, every decision. AI finally makes that aspiration economically viable, governed by people who decide what the system optimizes for, what it is forbidden to do, and what counts as proof. Judgment does not disappear. It moves upstream. The 10x Marketer is not trying to out-decide the machine. The 10x Marketer sets the goals, guardrails, and standards it follows—and writes its constitution.
The work starts smaller than the idea: Redesign one workflow end-to-end, start the decision log this quarter, and change what you promote for—reward the people who improve decisions, not the people who produce the most visible output.
AI will make every marketing organization faster. It will not make every marketing organization smarter. That difference will come down to judgment: what deserves attention, where capital creates value, what the system should optimize for, and what it should refuse to do.
That is the work of the 10x Marketer, and, increasingly, the work of marketing itself.
It’s not 10 times more output. It’s 10 times better decisions and an organization that keeps learning how to make them better.
Note: This content was created by Zeta Global.

