The 90/10 Framework: How to Balance AI Agents and Human Expertise in Your Marketing Strategy
After 18 years as a CMO, I've learned that the best marketing isn't about doing more — it's about knowing what only a human can do. The 90/10 Framework is how we think about that at Robynn AI.
Back in my CMO days at a company that will remain nameless to protect the innocent I was talking to the CEO, a brilliant engineer about hiring an Paid Campaign specialist. When I explained to him what this person is going to do, the CEO 's' question to me was, "That sounds like a Simpson character sitting in front of a giant board with dials. Why can't you just automate all of this with code?"
"All of this," of course, was just "performance marketing". If you now zoom out into the full marketing platform you will now see a sprawling, chaotic, beautiful mess of brand, demand gen, content, community, and a dozen other things we were trying to wrangle with a team that was perpetually understaffed and a stack that was perpetually over-tooled.
He'd just seen a demo of some new marketing automation platform that promised to be a "growth engine," and his engineer brain, quite logically, wanted to know why we couldn't just pour leads in one end and get revenue out the other, with no messy human intervention in between.
I didn't have a great answer at the time. I mumbled something about "brand nuance" and "creative judgment" and watched his eyes glaze over.
I think about that question a lot. It's the same question buzzing around the marketing world today, just with a new coat of paint. Except now, the word isn't "automation," it's "AI." Will AI take our jobs? Is the marketing department of the future just a single person and a very powerful GPU?
The conversation is dominated by two extremes: the utopian belief that AI will solve everything, and the dystopian fear that it will replace everyone.
They're both wrong.
After twenty-plus years of straddling the worlds of engineering and marketing — from shipping code to managing product strategy at Oracle to running marketing at multiple unicorns — I've learned that the most powerful solutions are rarely about "all or nothing." They're about balance. They're about, wait for it - relationships. This is especially true for an agentic marketing strategy. The future isn't a Skynet-style takeover or a Luddite-led rejection.
It's a framework I call the 90/10 split.
The philosophy is simple: 90% of the work in marketing — the repetitive, the scalable, the data-intensive — should be delegated to AI agents. The remaining 10% is where human craft, taste, and strategic judgment make all the difference. It's the partnership model, not a replacement one. And it's the only way to build a marketing function that is both ruthlessly efficient and deeply human.
With this framework now the utopian dream of marketing to a segment of 1 is finally achievable.
The 90% — What to Delegate to AI Marketing Agents
Let's start with the 90%. This is the work that, if we're being honest with ourselves, most marketers don't love doing. It's the grunt work. The endless data pulls, the mind-numbing report generation, the Sisyphean task of personalizing content for a thousand different segments. This is the work that burns out your best people and bogs down your boldest strategies.
I've managed marketing teams of fifty-plus people, and I can tell you that the vast majority of their time — the overwhelming majority — was spent on tasks that were:
- Repetitive: Anything you find yourself doing over and over in a spreadsheet. Pulling weekly performance reports. Updating campaign dashboards. Formatting content for different channels.
- Data-driven: Anything that requires synthesizing vast amounts of structured or unstructured data. Competitive analysis. Audience segmentation. Keyword research. Attribution modeling.
- Scalable: Anything where the process is more important than the individual instance. Generating content variations for A/B testing. Executing multi-channel campaign sequences. Monitoring brand mentions across the internet.
Think about the sheer volume of data a modern marketing team has to contend with. Website analytics, CRM data, social media engagement, ad performance, product usage signals, competitor moves, industry trends… it's an ocean. A human can only doggy-paddle on the surface. An AI agent can dive to the deepest trenches and map the entire seafloor.
This is where an agentic marketing framework shines. And I want to be precise about the word "agentic" here, because it matters. This isn't the old-school marketing automation that runs a pre-programmed drip sequence, batches and blasts 5 million emails and calls it a day. Agentic means the AI can reason, adapt, and take action based on context. Above all, it can follow a process you define. It's the difference between a thermostat and a person who understands that you're cold because you just came in from the rain, not because the room temperature changed.
At Robynn, we're building our platform around this principle. Our CMO agent, Rory, can perform SEO and competitive analysis at a scale no human team could match. Our Playbooks allow you to compose multi-agent workflows to execute entire campaigns — from research to outreach to performance monitoring — with drag-and-drop simplicity. The key is that these agents are grounded in a deep, living understanding of your brand — what we call the Brand Book. It's not just generating generic content; it's generating content that sounds like you, because it has learned you.
The goal of the 90% is to liberate your human marketers from the tyranny of the tactical. To let them off the assembly line so they can step into the design studio.
The 10% — Where Human Expertise Remains Irreplaceable
And that brings us to the 10%. If you automate 90% of the work, what's left?
Everything that matters.
This is the realm of taste, of creativity, of empathy, of strategic judgment. It's the work that can't be quantified in a spreadsheet or optimized with an algorithm. The 10% is where you tell your story — in a way no AI can tell it for you.
What does this look like in practice?
- Brand Strategy & Storytelling: An AI can tell you what resonates, but a human must decide why it matters. Defining your brand's soul, its point of view, its reason for being — that's a human task. It's the difference between a brand that has a mission statement and a brand that is on a mission.
- Creative Direction & Taste: An AI can generate a thousand images, but a human with taste must choose the one that feels right. Taste is a form of compressed experience. It's the intuitive understanding of what will connect with an audience, and it's built over years of paying attention to what works and, more importantly, what doesn't.
- High-Stakes Communication: Writing a crisis response, crafting a message to an early investor, or navigating a tricky partner conversation. These moments require a level of nuance, empathy, and social awareness that is, for now, uniquely human.
- Interpreting Nuanced Insights: An AI can tell you there's a correlation between two data points. A human needs to interpret whether that correlation is a meaningful insight or a statistical ghost. It's the ability to ask "So what?" and actually have a good answer.
- Building Relationships: True relationships — whether with customers, partners, or influencers — are built on trust and genuine connection. You can't automate authenticity. You can't scale a handshake.
The 10% is where you make the decisions that define your brand. It's where you take the data and insights from your AI agents and filter them through the lens of human experience. It's the art that complements the science. And it's why your AI needs a memory and brand context to be a true partner in this process — because without understanding your brand deeply, the 90% becomes generic noise.
How to Implement the 90/10 Framework
This all sounds good in theory, but how do you actually put it into practice? It's not about flipping a switch overnight. It's a gradual, deliberate process of shifting your team's mindset and workflows.
Step 1: Audit Your Workflows. Start by mapping out everything your marketing team does in a given week. I remember going through one of these exercises when I was running product marketing. Be brutally honest. What percentage of your time is spent on low-leverage, repetitive tasks? What are the things that only a human can do? Most teams I've worked with discover that 70–80% of their time is already spent on the 90% tasks — they just haven't categorized it that way.
Step 2: Define Human-in-the-Loop Checkpoints. Identify the key decision points in your workflows where human judgment is critical. This is where you'll build your 10% checkpoints. For example, an AI agent might generate fifty subject lines for an email campaign, but a human makes the final call on which one to use. An agent might draft a competitive analysis, but a human decides the strategic response.
Step 3: Start with a Pilot Project. Don't try to boil the ocean. Pick one well-defined workflow — like competitive monitoring, content ideation, or SEO optimization — and introduce an AI agent to handle the 90%. Measure the impact on both efficiency and quality. Learn what works before you scale.
Step 4: Measure and Iterate. The goal isn't just to save time; it's to produce better work. Are you shipping better campaigns? Are you getting deeper insights? Is your team happier and more engaged now that they're spending more time on creative strategy and less time on data wrangling? Use these metrics to refine your 90/10 split over time.
The Future is Partnership, Not Replacement
I never did convince that CEO to stop asking if we could automate everything. But the conversation did shift over time. At one point I remember ee started focusing on how we could use technology to free up our best people to do their best work.
That's the promise of the 90/10 framework. It's not about building a marketing machine that runs on its own. It's about building a marketing team that is amplified, not replaced, by intelligence. It's about creating a partnership between human and machine, where each does what it does best.
The future of marketing isn't about choosing between agents and humans. It's about realizing that the two together are more powerful than either one alone. It's a vision we're committed to, as you can read in the Robynn Manifesto.
Ready to see how your brand looks through this new lens? Start by letting our agents generate your Brand Book.
It's time to stop worrying about being replaced and start thinking about being amplified. 🖖
Continue exploring
View all
AI Agents for Marketing: A Comprehensive Guide
A comprehensive guide to ai marketing agents for differentiated marketings and steps to build a simple agent with n8n
Read more
The CMO’s Mission Impossible : Why Measuring Marketing Impact is Like Catching Smoke (And How to Do It Anyway)
A framework to measure marketing impact
Read more