AI in Marketing
Jul 9
Carl Pierre on where AI belongs on a marketing team
Carl Pierre on where AI belongs on a marketing team, which decisions it should support, and where human judgment still wins.
Carl Pierre on where AI belongs on a marketing team, which decisions it should support, and where human judgment still wins.
AI belongs inside the marketing team, but not at the center of the brand. The center is still judgment: what the company believes, what the customer actually needs, what the market is tired of hearing, and what the brand can credibly say better than anyone else. I am Carl Pierre, and the teams that understand that hierarchy will get more from AI than the teams that treat it as a shortcut around thinking.
The marketing team of the next few years will not be split between people who use AI and people who do not. It will be split between teams that design good decision systems and teams that paste outputs into campaigns because the tool made it easy.
Most people notice AI in the writing stage. That is understandable because writing output is visible. But the highest-value uses often happen before writing and after performance data comes back.
Before the draft, AI can help organize messy inputs. It can summarize customer research, group sales objections, identify patterns in reviews, compare competitor claims, and turn scattered notes into a usable brief. That work makes the human draft better because the thinking is better.
After the campaign, AI can help the team understand what changed. It can compare message variants, pull themes from comments or inquiries, find gaps between what the brand said and what the customer repeated, and help translate performance data into a clearer next test.
The middle stage, the actual creation of copy, still matters. But it should not be the only place AI shows up. If the input is weak, faster writing only spreads the weakness across more channels.
A lot of AI marketing work creates more noise. More emails, more variants, more captions, more dashboards, more half-finished ideas. The better use case is the opposite: reduce noise so the team can make cleaner decisions.
The useful question is not, “Can AI make this faster?” The useful question is, “Can AI make the team less confused?”
That changes the task list. Instead of asking AI for ten taglines, ask it to extract the five objections customers keep repeating. Instead of asking it for another content calendar, ask it to identify which themes are already overused in the category. Instead of asking it to write an ad from scratch, ask it to pressure-test the strategic claim before the budget goes live.
That is where the tool starts acting like an analyst, not a slot machine.
The cleanest operating model I have found is three lanes.
Lane one: research and synthesis. This is where AI can move fastest. Summaries, clustering, comparison, transcript analysis, review mining, competitor patterning, and first-pass audience research all belong here. The risk is manageable because the output informs judgment rather than touching the customer directly.
Lane two: assisted production. This includes briefs, outline drafts, email variants, ad copy options, social post drafts, landing page structures, and reporting summaries. Human review is required because the work carries the brand voice.
Lane three: governed automation. This is where the rules must be strict. Anything that sends a message, changes targeting, adjusts spend, scores a lead, or personalizes an experience should have a defined approval path and clear measurement. In hospitality and other trust-heavy categories, this lane should move slower than the vendors want it to.
The point is not to slow the team down. The point is to keep authority in the right place.
The Stanford 2025 AI Index Report makes the direction clear: AI is getting more capable, cheaper to access, and more embedded in work. That means AI fluency is no longer a technical specialty alone. It is becoming a management skill.
Marketing leaders need to understand prompts, model limits, data risk, workflow design, quality control, and measurement. They do not need to become engineers. They do need to know enough to stop bad systems from becoming normal systems.
That is the real risk. Not that AI makes one bad caption. The risk is that a company quietly builds a marketing process where nobody knows why the message is being said, who approved it, what data shaped it, or whether it still sounds like the brand.
Tools like Claude Enterprise point toward the enterprise version of the category: secure collaboration, knowledge integration, and workflow support. But the tool layer will keep changing. The management layer is what has to stay clear.
People often defend human judgment in sentimental terms. Creativity. Taste. Empathy. Those words are true, but they are not precise enough for how marketing teams actually work.
The human layer matters because someone has to own the tradeoffs. Faster versus better. Personalized versus invasive. On-trend versus indistinct. Efficient versus forgettable. Bold versus unsupported. No model owns those choices. The business does.
That is why AI strategy belongs close to brand strategy. Both are systems for deciding what the company should and should not say.
A team with weak brand judgment will use AI to become louder and more average. A team with strong brand judgment will use AI to see more clearly, test faster, and protect the voice that makes the brand worth choosing.
The shift worth making is to stop asking where AI can replace the marketing team and start asking where it can improve the team's judgment loop. Research should get sharper. Drafting should get easier. Measurement should get clearer. Governance should get stronger.
That is where AI belongs: not as the brand, not as the strategist, but as the operating layer that helps good marketers make better decisions faster. I write more about AI marketing, brand strategy, and hospitality marketing at carlpierre.com.
Carl Pierre is a performance marketing strategist based in the Washington DC metro area. He works in hospitality marketing and has built marketing systems across coworking, technology, and luxury hospitality. His work has appeared in Washingtonian, WAMU, Bloomberg, and Northern Virginia Magazine. More at carlpierre.com.
