AI inside the marketing function at 10M to 50M

A sober operator view

AI is changing marketing fast. Most of the noise is coming from people selling tools, not from people running businesses.

This essay is the operator view. Where AI is genuinely useful inside a B2B marketing function at 10M to 50M today. Where it is overhyped and where to wait. How to brief the team. How to think about the budget. And how to keep the company defensible while the technology moves.

We use AI inside every CMO Grow engagement. We also push back hard on the hype that comes through the marketing trade press every week. This is what we tell CEOs who ask, what should we do about AI in marketing.

Where AI is genuinely useful today

Four places, in our experience.

First, research and synthesis. The team can read three competitor sites, four buyer interviews, and a pile of analyst reports in an hour with an AI assistant. The AI does not replace the analysis. It compresses the time to get to the analysis. A senior marketer who used to spend a week on a positioning research pass now spends a day. That is real leverage.

Second, drafting. Drafting first versions of emails, ad copy, sales pages, blog posts, and one pagers. The AI gets you to a 70 percent draft fast. A senior marketer takes it to 100 percent. The result is shipped faster, with more variations tested, without sacrificing voice. The trick is that the senior marketer still has to be senior. A junior team using AI to draft from scratch produces fluent garbage.

Third, structured data work. Cleaning a CRM. Categorizing inbound leads. Building lookalike lists. Summarizing customer calls. Pulling themes from open ended survey responses. This is unglamorous work that ate junior hours for years. AI does it in minutes.

Fourth, internal documentation. The operating cadence, the playbooks, the briefing notes, the handover docs. The team that uses AI to document the way it operates ends up with a more durable system, because the documentation actually gets written. Without AI, most marketing teams skip it.

Where AI is overhyped right now

Three places, in our experience.

Autonomous campaigns. The vendors will sell you a tool that promises to run your demand engine end to end with no human in the loop. That is not real today. The tool produces output. The output ranges from bad to mediocre. A senior marketer who reviews and tunes the output gets a useful result. A team that turns it on and walks away gets exposure.

Hyperpersonalization at scale. The vendors will sell you a tool that promises to personalize every touchpoint in real time based on signals. The math sounds right. The execution is mostly cosmetic. The buyer notices when personalization is wrong far more than they notice when it is right. The downside of getting it wrong is bigger than the upside of getting it right, for most companies at this stage.

AI generated thought leadership. Long form content written by AI sounds like long form content written by AI. Sophisticated buyers notice. Trust gets damaged faster than the content team realizes. Use AI for research and drafting. Do not publish AI generated thought leadership without senior human authorship.

How to brief the team

The brief is simple. Three rules.

Rule one. Senior eyes on every published piece of work. AI is a force multiplier on senior judgment. It is not a replacement for senior judgment. If the company does not have senior marketing leadership, the AI output ships unreviewed and the brand pays for it.

Rule two. Documented use, every time. Whichever AI tool the team uses, the team logs the prompt, the output, the human edits, and the published version. Two reasons. Training data. And, more importantly, defensibility if a regulator or a customer asks how a claim was generated.

Rule three. No customer data in public AI tools. Personal data, customer lists, contract terms, financial data, and product roadmaps do not get pasted into ChatGPT, Claude, or any consumer AI tool. The team uses enterprise grade tools with the right data residency and contract terms for anything that touches customer data.

These three rules are simple. Most marketing teams break at least one of them within 90 days of starting to use AI. The cost of that break ranges from embarrassing to existential.

How to think about the budget

AI is not a line item. It is a multiplier on the senior marketing team that already exists.

Two budget shifts we see working at 10M to 50M.

Shift one. Move some junior production budget into senior strategy budget. The AI does the production work. The senior person does the strategy work. Net is the same or lower total cost, with better quality output.

Shift two. Add a small enterprise AI tooling budget. Two to five hundred dollars per seat per month for the marketing team's primary AI stack is normal at this stage. Bigger budgets are usually a sign of tooling sprawl, not strategy.

CEOs sometimes ask whether they should hire an AI marketing specialist. The honest answer is, not yet. The right move at the 10M to 50M stage is to make every existing marketer AI fluent. Specialists make sense at scale beyond this band.

Where it goes from here

AI in marketing is going to keep getting more capable over the next 24 months. The defensible move is not to bet on a specific tool. The defensible move is to build a marketing function that can adopt new capabilities without losing the brand or the customer trust that took years to earn.

That is a leadership question more than a technology question. The marketing leader who can keep the team disciplined about AI use is the same marketing leader who can keep the team disciplined about anything else. The technology rewards the function that is already running well. It exposes the function that is not.

What to do next

If you want a clearer read on how AI fits into your operating system, take the Growth Assessment Scorecard. Twelve questions. Six minutes. A personalized read across strategy, leadership, and execution.

If you would rather just talk it through, book a 30 minute discovery call. We will give you the operator view on AI in your specific situation, with no tool pitch.

---

Chris Lundell is the founder of CMO Grow. He is a three-time CEO across enterprise software and residential solar, and currently serves as Chief Compliance Officer and a Board Member at SunPower.