Marketers Are Becoming AI Builders. Now We Need to Learn How to Build.

There is no shortage of opinions about what AI is going to do to marketing.

Depending on your LinkedIn feed, AI is either going to eliminate the marketing department, give every marketer superpowers, destroy search, make agencies irrelevant, or allow one highly motivated person with a Claude account to run a Fortune 500 company from a coffee shop.

Some of this will age about as well as the metaverse. But underneath the hype and fear mongering is something very real: AI is a massive change for marketing, and I don't think most of us fully understand where it's going yet.

I'll be a little “smarky” along the way. The AI conversation could use less breathlessness and more focus on what marketers should actually do.

The timing is intentional. Today we're launching BlueRock for Builders around a change I increasingly see firsthand: marketers and other knowledge workers aren't just using AI anymore. We're becoming builders.

AI gives marketers leverage. It doesn't give them taste.

We are going to produce more with smaller teams. Research will happen faster. Campaigns will be improved continuously. Personalization that would have been economically ridiculous a few years ago will become commonplace and even a little scary.

But leverage works in both directions. AI can make excellent marketing faster. It can also make terrible marketing at previously unimaginable scale. And we are already seeing plenty of both.

Give it generic positioning, a fuzzy understanding of your customer and a mediocre campaign idea, and it will happily produce 47 versions by lunch. Congratulations. Your content engine is humming.

That is why I think the fundamentals of marketing become more important, not less. If I had to boil them down:

• Know your customer. Understand what they actually care about, not what you hope they care about.

• Know what you stand for. Great positioning and a differentiated point of view become more valuable when everyone can generate content.

• Have ideas worth paying attention to. More output does not equal more impact.

• Develop judgment. Knowing what not to make may become one of the most valuable skills on the team.

When production becomes nearly unlimited, production stops being the scarce resource. Taste, judgment and ideas still are.

And there is another scarce resource that AI is starting to change: our ability to actually build the things we need.

I used to shop for software. Now I'd build it.

Before joining BlueRock, most of my AI experience was similar to what I saw across marketing: research, writing, analysis, ideation and summarization. Those capabilities alone had already changed how I worked and made me considerably more productive.

What changed at BlueRock was getting deeper into building agentic systems myself. That's when I started to understand the Citizen Developer idea in a very practical way: marketers like me can increasingly build the software we used to buy.

Over my career, I've probably been involved in tens of millions of dollars of marketing technology purchases. CRM, marketing automation, analytics, intent data, content systems, ABM platforms, social tools and plenty of things whose category names I've thankfully forgotten.

The process was usually some variation of the same thing. We had a problem, found software that mostly solved it, bought it, integrated it, trained people and adapted our process around whatever the software was designed to do.

Today, I increasingly look at marketing problems differently. Instead of automatically searching for a product that gets me 70% of the way there, I can build around the exact problem I have.

That changes the equation. Instead of buying a platform built for thousands of companies and adapting our process around it, marketers can increasingly build systems around how their own teams actually work. Research can run continuously. Insights can surface when something changes. Work can move between specialized agents instead of waiting for the next human handoff.

We're not just getting better marketing software. We're gaining the ability to create it.

Agentic AI Is More Than Automation

Here is where the smark stops for a minute, because this distinction matters.

For 25 years, most marketing technology followed rules we defined. If X happens, do Y. Send this email. Update this field. Trigger this campaign.

Agents are different. Give them an objective and they can reason about what to do next, use tools, work with other agents and take action without someone directing every step.

That creates new requirements. Agents need somewhere to run when your laptop is closed, persistent context to do their jobs, boundaries around what they can do, and visibility into their actions and costs.

The more autonomy we give AI, the more important the environment around it becomes.

So yes, marketers can suddenly build much more themselves. The next question is: where should they build it?

Let your Citizen Developers build. Securely.

Right now, companies often give aspiring AI builders two lousy choices: wait for IT and security to create an approved path, or go full YOLO with personal accounts, credit cards and laptops.

One kills momentum. The other eventually gives your CISO a migraine.

We can do better than that.

That's a big part of why we're launching BlueRock for Builders. The idea is simple: enable this new generation of Citizen Developers and AI builders instead of blocking them.

Builders stay in tools like Claude Code or Cursor, while BlueRock gives their agentic work a persistent AI workspace with continuous compute, context, visibility and controls built in.

We've also created BlueRock Learn to teach people how to build. Start with an agent, give it memory and skills, assemble agent teams, then put useful work on a schedule and let it run.

You don't have to become a traditional software developer. You do have to start thinking like a builder. And I think we're going to see a lot more of them inside marketing organizations.

Go Build Something, Marketer!

I don't know exactly what the marketing organization looks like five years from now. Anyone who tells you they do probably has a keynote to sell you.

But I'm convinced the marketers who thrive will combine great marketing judgment with the ability to build with AI. They'll understand customers, positioning and ideas, then turn that understanding into systems that actually do work.

So here's my pitch: stop reading about agentic AI for an afternoon. Go build something. Pick a real problem. Give an agent a job and context. Connect it to another agent. Put something useful on a schedule. See what happens.

And if you want a place to start, try BlueRock for Builders. Then tell us what worked, what didn't and where you got stuck.

The people who know the problem can increasingly build the solution. That's a lot more interesting than another 10% productivity gain.