Marketing Tech Founder And CMO Stresses The Importance Of Context Before Integrating AI
Learn how a solo marketer uses AI and context engineering to scale campaigns, automate workflows, and compete with larger teams, with practical insights on adoption and execution.
AI Leverage: Jean Bonnenfant uses AI to amplify his one-person marketing efforts at startup Tinct.ai.
Context Engineering: Building a robust context layer is crucial for effective AI-driven marketing strategies.
Outbound Automation: AI enhances outbound marketing with personalized sequences and dynamic landing page experiences.
Retention Redesign: AI can revolutionize customer retention by enabling personalized communication and proactive outreach.
Human-AI Balance: Successful marketing integrates AI efficiency with human oversight to maintain strategic relevance.
Jean Bonnenfant has a history of leading growth and marketing at B2B tech companies, and he's currently the cofounder and CMO at the early-stage startup, Tinct.ai. As a marketing department of one, AI is the only way he can compete. And he's leveraging it to produce more than most small teams.
We sat down with him to figure out how he's doing it. He said it's all about context engineering.
One AI-enabled Marketer Can Outperform A Team
I've been in marketing for almost 10 years. I started as a marketing manager, focusing mostly on content with some growth work. That led me to join Growth Tribe, Europe's first academy for growth hacking and AI, and my relationship with AI began there, long before today's LLM boom.
Want more from The CMO?
Sign up for a free membership to complete reading this article:
Have an account? Log In
At Growth Tribe, we ran experiments with machine learning tools. We worked with companies like Dataiku, a French ML company, integrating AI into client projects when data volumes justified it. This included clustering, predictive modeling, and pattern recognition. It wasn't flashy, but it worked, and it gave me a solid foundation for understanding what AI can do versus what it promises.
After that, I worked in FinTech and crypto, two industries where data and speed matter significantly. Then, just before founding Tinct AI, I worked at Lleverage, an AI automation platform, operating at real scale with AI tools. That experience was the final piece that gave me the confidence to build something of my own.
Today, I'm the cofounder and CMO of Tinct AI. We build AI-powered personalization for B2B landing pages, specifically for ABM campaigns. Every company you target sees a version of your landing page tailored to them, identified via IP recognition. It's the kind of product I wish had existed when I was running campaigns earlier in my career.
Our goal is to scale without increasing headcount, which is only possible because I work with AI. So, what I do now looks very different from traditional marketing org structures. I run all channels daily, collaborating with AI workflows. ICP definition, messaging, CRM integration and enrichment, outbound sequences, content production — it's all me, but it's never just me. My bet is that one person operating with the right AI infrastructure can outperform a team of five working without it.
How Context Engineering Is a Vital Marketing Skill
When LLMs first took off, most people used them reactively. Users had a task, opened a chat, and got something out. "Write me an email sequence" or "Draft me an article." This approach was punctual, transactional, and ultimately, limited.
I stopped thinking about AI as a task executor and started viewing it as an infrastructure layer. Context forms the foundation of that infrastructure.
Context engineering is now one of the most important skills in marketing. This includes tools like Claude Projects, system prompts, skills, and memory layers — anything that provides an AI with deep, structured context about your ICP, positioning, USPs, competitive landscape, and tone. When the context is right, everything built upon it is also right. The output quality compounds. If that layer is vague, generic, or outdated, everything AI produces on top of it will be too.
Every marketing leader starting their AI journey should first obsess over that context layer. Make it specific. Make it accurate. Make it yours. And then make sure every tool, every workflow, every automated output is dipped into it.
If I had known to invest in the context layer first, I would have avoided weeks of inconsistent output and rework.
So, the biggest change I've made in the last year has been investing significant time up front to build that context layer across all my systems, workflows, and platforms. Every time our positioning evolves, or we learn something new about our ideal customer, I update the context across the board.
As a result, I can produce and distribute assets at a speed and consistency that would normally require a full team. The foundation does the heavy lifting. What I build on top of it almost takes care of itself.
When the context is right, everything built upon it is also right. The output quality compounds. If that layer is vague, generic, or outdated, everything AI produces on top of it will be too.
How To Leverage AI In Outbound Workflows
Here's how I run outbound now.
The process starts with list building and enrichment. I pull a list of target accounts from our ICP definition and use AI-powered enrichment to add firmographic data, tech stack signals, and relevant context about each company. That data feeds directly into lemlist to run outbound, while Attio automatically enriches and updates as new signals come in.
Next, I move into messaging. Using a Claude Project loaded with our full ICP context, positioning, and tone guidelines, I generate personalized outbound sequences for each segment. This involves not just swapping a first name, but actual message variation based on the company's profile, their likely pain points, and where they sit in our target account list.
When prospects click through to our landing page, it is automatically personalized to their company, identified through IP recognition. The messaging, social proof, and framing all adapt to who they are. The personalization that started in the outbound email doesn't stop at the click; it continues through the entire experience.
Replies and intent signals flow back into Attio, and then I step in personally. Everything before that moment is AI. Everything from that point is me.
I am part of the workflow at two points: the setup and the conversation.
Why Retention Strategies Need An AI Revamp
Top of funnel gets all the AI love because it's the most visible part of the job. Ads, content, campaigns. It's easy to point to and easy to measure. And AI has already started reshaping it significantly.
But retention is where most companies are still operating like it's ten years ago. Manual touchpoints, broad segmentation, and generic nurture sequences that treat every existing customer the same, regardless of how they behave or what they need.
That's the stage most in need of redesign, because AI makes something possible there that wasn't before: treating every single customer as an individual, at scale, without a team of ten to manage it. Personalized lifecycle communication, dynamic content based on usage patterns, proactive outreach triggered by real signals rather than arbitrary timelines.
The irony is that retention is where the money is. Acquiring a new customer costs five to ten times more than keeping an existing one. And yet most marketing orgs still overwhelmingly point AI investment and creative energy at the top of the funnel.
That's the redesign opportunity. And most companies haven't touched it yet.
How AI's Benefit Is Also Its Greatest Risk
In the last two months, I've shipped what a traditional setup would have realistically taken a year to complete. Full outbound infrastructure, CRM enrichment, content production across channels, design partner onboarding, and messaging frameworks. I completed all of it solo, with AI as my copilot.
There's a narrative that suggests you can work half an hour a day if you use AI. That's not what happened for me. Instead, the ceiling on what I could accomplish moved dramatically. I didn't do less work; we did exponentially more.
Beyond volume, the error rate on certain tasks also dropped noticeably. Data entry, research, structured outputs. AI is more consistent than a tired human at the end of a long day.
With that said, quality control at speed is a challenge. When you're moving fast and producing a lot, things risk slipping through that a slower, more manual process would have caught. You must stay disciplined, preventing volume from becoming an excuse for lower standards.
But net-net? The productivity delta is real and significant. AI didn't give me leisure time; it gave me leverage.
How AI Is Disrupting Copywriting
I'm a copywriter by nature. Writing has always been my thing, and for a long time, I believed it was the last skill AI would ever credibly replicate. You can automate data, automate distribution, automate targeting. But voice? Nuance? The ability to make someone feel something with a sentence? I was convinced that was human territory.
And I still think that's true for literature. I haven't seen a novel or a book written by AI that moves me. That frontier still belongs to humans.
But for business writing? Month after month, I watch it get better. And with context engineering, the gap has closed faster than I ever expected. Give an AI a structured ICP, a detailed style guide, real brand examples, and enough context to work from, and the output can genuinely carry a brand's voice. Not just passable. Good.
At a previous company, I built an entire blog using a Claude Project, which I trained on a very specific system prompt and a long, detailed style document. This was not a few bullet points about tone, but an actual, multi-page document built around writing that I genuinely loved reading and wanted to emulate.
The output wasn't generic AI content. It had a real, consistent voice. If you know writing, you wouldn't mistake it for something else. It had a rhythm, a perspective, and a way of framing ideas that felt intentional.
So, the assumption I held for years, that AI would always struggle with the craft of writing — it's gone. And it's beautiful to watch. Even when it disrupts something I thought was mine.
How AI And Human Roles Blend In Marketing Decisions
In practice, there isn't a clean split between what humans should do and what AI should do. It's never a binary. It's a spectrum.
AI informs or powers almost all of our decisions. Outbound sequences, content, CRM enrichment, ICP refinement, messaging. AI is involved at every stage. But humans are not out of the picture.
Think of it this way: AI handles the volume, the speed, and the first draft of everything. A human is always in the loop to catch edge cases, apply judgment, and ensure what goes out makes sense in context. This human layer isn't a specific role or task. It's woven into every workflow.
Discernment remains human. Knowing when something is technically correct but strategically wrong. Sensing when a message will land or fall flat. Seeing when a trend is worth jumping on versus when it's going to age badly in two weeks. Understanding the nuance of a conversation with a potential partner that no prompt can fully capture. Those are the moments I step in, and they happen constantly.
So, it's less about AI versus human activities and more about ensuring a human hand always guides the direction, even when AI does most of the work.
Discernment remains human. Knowing when something is technically correct but strategically wrong. Sensing when a message will land or fall flat. Seeing when a trend is worth jumping on versus when it’s going to age badly in two weeks.
Why AI Adoption Is More Important Than Tooling
The tools are already incredible. In fact, they're ahead of our collective ability to use them well. And that gap between what AI can do and what most people get out of it is not a tooling problem. It's a human problem.
What makes it particularly interesting is that even the people actively using AI every day feel behind. I feel behind. A new model, capability, or way of working always emerges, often just last week, and I haven't absorbed it yet. That's my reality, and I am deeply curious and deliberately invest time in staying close to this space.
When I consider someone less curious, less plugged in, less motivated to experiment, the gap is enormous. Not because the tools are hard to use, but because adoption requires a mindset shift that no software update delivers.
The biggest unlock in any marketing org isn't finding a better tool. It's upskilling your people to think differently about how they work. What tasks they delegate to AI, how they structure their prompts, and how they build and maintain context. That's the capability that compounds, and most organizations underinvest in it.
The Case For Marketing Leaders Breaking Things
My advice is simple: Experiment. More than you think you need to.
This is not a moment to delegate your AI curiosity to someone on your team and wait for a report. You need to get your hands dirty yourself. Build things. Break things. Try workflows that might not work. The only way to develop genuine judgment about what AI can and can't do is through direct experience, not observation.
Leaders tend to stay at the strategic layer, letting others do the operational work. That instinct needs to go right now. The gap between leaders actively building with AI and those just talking about it is widening fast. It shows in the quality of their decisions.
Being a leader doesn't exempt you from getting in the weeds. If anything, this moment demands the opposite. The more you experiment personally, the better your instincts become, the better your team becomes, and the better your strategy becomes.
So experiment. Build. Stay curious. And don't wait until you feel ready, because that moment won't come. The only way through it is through it.
Follow Along
You can follow Jean Bonnenfant's journey on LinkedIn as he single-handedly grows Tinct.ai.
More expert interviews to come, so sign up with a free account to follow along on The CMO Club!
