CMO Says Most Marketing Teams Are Producing Mediocre Work Faster With AI
Discover how a CMO uses AI to redesign marketing workflows, accelerate decision-making, and build scalable growth systems without sacrificing strategic thinking or creativity.
Samantha Leal
CMO of Aliwell
AI Workflows: Samantha Leal implements AI in marketing, redesigning workflows for growth, but faces varied success.
Resource Constraints: AI enables marketing leaders to build strategic systems that small teams couldn't normally manage.
Creative Limits: AI struggles to produce high-quality visual content, affecting performance in specialized markets like orthodontics.
Execution Bottleneck: AI shifts marketing bottlenecks from execution to strategic judgment, requiring more human decision-making.
Strategic Clarity: Effective AI use in marketing demands strong data inputs and clear strategic direction to avoid mediocrity.
Samantha Leal is the CMO of Aliwell, where she is implementing AI workflows, some more successfully than others.
We interviewed Samantha to learn what worked, what didn’t, and why. Here's what she told us.
Going from zero to scale
Hey! I'm Samantha Leal. I’ve spent the last 10+ years helping startups go from zero to scale, usually stepping in when growth looks good on the surface but breaks under pressure.
Want more from The CMO?
Sign up for a free membership to complete reading this article:
Have an account? Log In
I work across the full growth system, from positioning and messaging to acquisition, conversion, and the customer journey. I partner closely with founders to bring structure to how decisions get made and how growth compounds.
Currently, I'm CMO at Aliwell, where I lead the marketing department across a multi-channel growth engine focused on acquiring and activating orthodontists. We operate across paid media, education, events, content, and sales enablement, but we primarily focus on activation and revenue.
We’re selling into a clinical audience, so the challenge isn’t just attention; it’s trust, education, and behavior change. That makes the system more complex. Tighter alignment with sales, product, and customer experience is critical.
Shifting focus to AI
What’s changed for me recently is how I approach all of this with AI.
I’ve stopped thinking in marketing roles and started thinking in workflows. Most of what drives growth — testing, creative iteration, reporting, lead qualification, etc. — is repeatable, but still manual and fragmented.
AI is the lever to redesign those workflows, speed up feedback loops, and make better decisions with greater clarity. That’s the shift I’m focused on now.
How AI changes what resource-constrained marketing leaders can build
In the last year, I used AI for some heavy lifts:
- I architected a full CRM migration that would normally require a consulting engagement.
- I designed a multi-tier education program from scratch.
- I built a customer activation system grounded in qualitative research — interviews, surveys, behavioral data — that uncovered why some customers stall after their first purchase.
These aren't tasks I sped up; they're strategic systems a small team couldn't normally ship. AI has changed what's possible for a resource-constrained marketing leader to build.
AI has changed what’s possible for a resource-constrained marketing leader to build.
How AI can be used for stakeholder reports
I use AI to generate a weekly report — a deliverable that drives company decisions for the week.
Inputs include meeting transcripts from the previous week (pulled from Granola), Meta Ads performance CSVs, ERP production data, social media analytics, and CRM pipeline updates. These five sources do not communicate.
Here's the workflow:
- I feed the raw transcripts into Claude and extract decisions, blockers, and departmental action items.
- I upload ad performance data and ask for week-over-week and month-over-month analysis, flagging anomalies.
- I analyze production numbers — regional cases fabricated, compared against monthly targets, highlighting variance.
- I synthesize all this into a single structured report with one section per function, key metrics with context, and — most importantly — a "decisions needed" section at the top.
The output is a formatted document that the leadership team can scan in five minutes and act on immediately. It contains no raw data or "here's what happened." Every number has a benchmark, relevant context, and a recommendation.
Where humans must maintain ownership
So, I rely on AI to power anything that benefits from speed, pattern recognition, and synthesis across large inputs.
That includes analyzing customer data across sources, clustering feedback from sales calls and interviews, generating and iterating on creative angles, and identifying patterns in campaign performance faster than a human could.
It’s also become a core part of how I structure workflows, from lead qualification to follow-ups to reporting. AI helps reduce manual work and surfaces insights earlier, so we’re not operating on lagging information.
Where I don’t rely on AI is in defining strategy, positioning, and what actually matters.
Decisions like what we say, who we’re for, what trade-offs we make, and where we focus are still human. Those require context, taste, and judgment, especially in complex markets where the right answer isn’t obvious.
AI can give you options, but it can’t tell you what’s worth doing.
Why AI shifts the bottleneck from execution to judgment
The biggest upside has been speed and scope. We’ve reduced time on analysis and synthesis by roughly 60–70%. Things like clustering customer feedback, extracting insights from calls, and generating testable creative angles — all of which used to take weeks — now happen in days.
That's translated into more testing volume and faster feedback loops. We're able to iterate on messaging and campaigns much more quickly, which has improved decision-making and reduced wasted spend on ideas that don't work.
On the operational side, we’ve also been able to build systems, like CRM workflows, onboarding journeys, and reporting layers, that would normally require a larger team or external support.
But real downsides exist. For example, AI can create a false sense of confidence. It’s very good at producing outputs that look right but aren’t grounded in reality. If you don’t anchor it in real customer data, you end up with generic messaging and poor decisions, just faster.
It also requires more judgment, not less. The bottleneck shifts from execution to deciding what is worth doing. Teams that don’t have that clarity can end up generating a lot of noise.
Why AI creative is a risk in specialized markets
I've been disappointed by AI's impact on creative. I expected AI to transform how we produce visual content — for example, ads, social posts, brand materials. It hasn't. Not even close.
Every AI-generated image still looks like an AI-generated image. For a brand selling to specialists — in our case, orthodontists who are highly educated and deeply skeptical — anything that feels generic or synthetic kills trust instantly.
We tested AI-generated visuals for ads and social content. Performance dropped immediately. Our audience noticed the quality gap between those visuals and what a human designer produces with real clinical photography.
Why content workflow sophistication must match capabilities
Early on, I used AI to design a content operations system — editorial calendars, approval workflows, and content briefs with detailed frameworks.
The system was strategically sound and beautifully structured, but it had a problem.
I handed a sophisticated system to a team member who wasn't performing at that level. Instead of the system raising her output, it created a gap. She couldn't execute against it.
I wasted weeks managing the gap between the system's demands and the person's capabilities.
Now, I ask, "Who will run this on Monday?" before I build anything.
The system was strategically sound and beautifully structured, but it had a problem. I handed a sophisticated system to a team member who wasn’t performing at that level. Instead of the system raising her output, it created a gap.
How to ensure AI messaging is not a hindrance
AI doesn’t fix bad systems; it amplifies them. More output, same problems. AI works best with strong inputs, clear customer understanding, real data, and sharp thinking.
So the question isn’t “how do we use AI?” It’s “What system are we plugging it into?”
A good example is messaging. In my first pilot, we moved fast but didn’t base it sufficiently on real customer data. The output looked strong, but it was generic. It didn't convert.
So, fix the inputs before using AI. That means consolidating data across the CRM, sales, and customer interactions — including sales calls, objections, and interviews — to identify the exact pains, language, and moments that drive decisions.
Then, I define clear angles and our core message. Once that’s solid, I layer AI on top. It helps expand, test, and iterate on those angles quickly.
From there, it should be a continuous loop. You constantly pull inputs from sales, customer feedback, and performance, and iterate in real time.
Why workflows need to be broken down and augmented
Done right, taste is no longer something you either have or hire for.
It can be shaped through process. With the right inputs, constraints, and iteration, you can systematically improve the quality of thinking and output.
Messaging, creative, research, even strategy. They're all workflows that can be broken down. Each step can then be augmented.
Why Claude is a must-have for marketers
Claude is my go-to tool, hands down. It's a layer that connects everything and acts like a great thinking and sparring partner.
I'm most excited about Claude Code and Claude Cowork because they move beyond chat into work environments like coding, file management, and task automation.
I haven't tried Claude Code with Cursor yet, but that's next!
How AI is changing teams
As far as team structure, we rely less on narrow specialists and more on people who can think across the system. People who understand the customer, can frame problems well, and know how to use AI to execute.
The execution work — production, analysis, and iteration — is no longer the bottleneck, so the value shifts. Now, I look for judgment, taste, and the ability to design workflows. People who can go from insight to action quickly, not just own a function.
It’s also changed how we collaborate. The lines between roles are less rigid. Marketing, sales, and customer experience are more connected because they share workflows.
Why CMOs must focus on clarity
Most marketing teams using AI are producing more mediocre work faster.
So, here's my advice: Focus on data inputs and strategic clarity.
When AI is used in marketing without strong data inputs and strategic clarity, it’s just a content pollution machine.
Follow along
You can follow along with Samantha's journey on LinkedIn!
And stay tuned for more expert interviews on The CMO Club!
