🧠 I was reading Function Health’s Growth Marketing Manager role and had the reaction I keep having lately: this looks like a normal marketing job until you notice how much of the work assumes AI as basic operating literacy.
The title is familiar. The job description is not.
We’re seeking a highly creative, AI-first Growth Marketing Manager to drive member activation across our employer (B2B2C) channel. This is a hands-on builder role for someone who is equally strong in strategy and execution—someone who can design growth systems, run campaigns, and use AI to fundamentally rethink how activation is done.
This is the part that stood out to me.
Lead with an AI-First Growth Approach
- Use AI tools to generate campaign ideas, messaging, segmentation strategies, and experimentation roadmaps.
- Build and optimize AI-powered workflows for campaign creation, testing, and iteration.
- Continuously identify ways to replace manual processes with AI-driven systems to increase speed and performance.
- Bring new, creative approaches to growth that go beyond traditional lifecycle marketing.
Function Health is not hiring someone to occasionally prompt ChatGPT for copy ideas. The role describes a marketer who can use AI across campaign ideas, messaging, segmentation, experimentation, workflow design, testing, and iteration.
That is a different operating model.
The job still sits under growth marketing, but the work is moving closer to… systems design? Marketing ops? We are used to strong marketers based on judgment, taste, customer understanding, and product instinct. The new layer is the ability to turn that judgment into workflows other people can reuse.
The Function Health post useful is a public example of something I think many teams are about to feel more directly: AI proficiency is becoming core part of the growth marketer’s job, not a skill like “Salesforce proficiency”.
The marketer is becoming the system designer
When I look at a marketing task now, I am usually not asking, “Can AI do this faster?”
Of course it can!
The issue is not automating the individual tasks, that’s easy. It’s chaining them. The friction created by the small handoffs:
- Someone rewrites the campaign brief from memory
- Someone pulls an audience list manually
- Someone recreates a nurture sequence from an old email
- Someone waits for data before deciding what to test next
- Someone knows the brand rules, but they haven’t been updated in the guidelines
- Someone has to remember what worked last quarter
A chatbot on top of a messy process gives you faster mess. A defined workflow with good context, sources of truth, and quality gates can give you a real speed advantage.
Auditing marketing workflows for AI readiness
If I was the AI-first Growth Marketer in question, I would start with one workflow. There are many, often undocumented:
Pick something specific:
- Lifecycle campaign build
- Audience segmentation
- Landing page testing
- Nurture sequence creation
- Onboarding, renewal, or reactivation comms
Then map what actually happens today. Not the clean version in someone’s head. The real version:
- Where does the work start from a blank page?
- Where does one person become the bottleneck?
- Where do approvals slow everything down?
- Where does quality depend on taste that has never been documented?
- Where do the learnings disappear after the campaign ends?
This is usually where the opportunity shows up to “agentify” the workflow, not just a task.
AI agents are useful when the job is defined
We know by now AI needs context. If the instruction is “write lifecycle copy,” the output will probably be generic. The AI has too much room to guess.
If the instruction is “turn this approved campaign brief into three email variants using our lifecycle template, source notes, ICP language, and voice guidelines,” the team has something it can review.
That is the difference between playing with AI and delegating work to an AI agent.
The agent needs:
- A clear job
- The right source material
- A definition of done
- Brand and compliance constraints
- A human review point
- A way to capture what worked
Without those pieces, the team spends the saved time fixing the output.
Wins should leave artifacts behind
This is the part that’s most promising, and most foreign to marketing teams, except the large, established ones.
A good campaign should not end as a Slack celebration and a line in a dashboard. It should leave something behind:
- A reusable brief
- A segmentation rule
- A tested messaging angle
- A QA checklist
- A comms calendar
- A playbook for the next version
Function Health uses language like “repeatable activation playbooks” and “systemized growth motions.” That is the phrase I would pay attention to.
The win is not only that one campaign worked. The win is that the next campaign starts from a better place.
Quality needs an owner
The reason teams get nervous about AI in marketing is not irrational. The failure modes are real.
AI can produce:
- Confident but unsupported claims
- Off-brand phrasing
- Wrong numbers
- Polished paragraphs that say very little
- Content that sounds like every other company talking about AI
That last one matters more than people think. Generic content erodes trust because readers can feel when nobody specific is behind it.
So the workflow needs quality gates:
- Source-linked drafts
- Clear claims review
- Voice and tone review
- Human approval for anything public
- A way to mark the final version as approved
At Eldur Studio, this is exactly why we heavily use Notion verification on source material, and voice guidelines. It forces me to review everything every 90 days and… guess what - positioning often shifts a bit. That voice and tone need to be fine-tuned after the latest LinkedIn post fiasco, or whatever has happened.
The New Growth Marketing Job
I wish Function Health the best in finding the right candidate. Maybe it’s someone internal who starts from one operational target:
Cut campaign iteration cycle time in half without adding headcount.
Then picks one workflow and rebuild it:
- Map the real process
- Document the sources of truth
- Add AI where it reduces handoffs or blank-page work
- Add quality gates before anything ships
- Turn the workflow into a playbook
That is where the “AI-first”marketer becomes valuable. Not because they use the newest model/skill/tool, but because they can make the work more repeatable, measurable, and durable.
The job title may still say Growth Marketing Manager. The job is very different now.
— Veronica