The Rise of the Marketing Engineer

I spent 12 years as a marketing generalist.

I was dialed in on strategy, content, and angles that would land with a buyer, and on a bias to ship something before it felt finished. Where I struggled was on the technical side. Anything that involved custom setup, integrations, or building tools usually meant waiting on engineering or product to free up.

That’s not how the job works anymore. The last 18 months changed what one marketer can build, and the gap between thinking it and shipping it has gotten very small.

Why generalists were undervalued for 15 years

For most of my career, marketing teams were built around two roles.

Specialists owned a single channel: paid, SEO, lifecycle, product marketing, each one a deep skill with a narrow lane. Generalists were everywhere at once, useful across most of it but rarely the expert on any of it. They were also the people stitching everyone else’s work together so the campaign actually launched on time.

The specialist model works when budgets are big enough. The trouble is that specialists are expensive, they don’t cross-pollinate easily, and you usually still need a generalist on the team to connect it all. The generalist model works when budgets are tighter, and the work has to ship anyway, which is where most marketing teams actually live.

The catch was always execution. A good generalist could plan smart work, but the technical backbone of modern marketing (integrations, reporting infrastructure, custom tools, data pipelines) needed someone with engineering chops. The best ideas tended to die quietly on a whiteboard.

That changed when AI tools got good enough for non-technical people to build with.

What I can actually do now that I couldn’t 18 months ago

I don’t code. I never have.

But you don’t need to anymore. With Claude Code and a working understanding of how to plug things into the tools you already use, you can build most of what a small marketing org needs:

• A market research system
• A content creation pipeline
• A paid ad creator and auditor
• An outbound email sequencer
• A weekly performance report
• Alerts for when something breaks

Two years ago, building those would have meant hiring 4-5 people and waiting a quarter for it to come together. Today, it’s a few weeks of focused work for one person who knows what they’re trying to build. I broke down what these systems actually look like in your terminal is the new marketing department.

The 67-point gap

The number that keeps coming up in my head is from Scott Brinker and Frans Riemersma’s 2026 Martech report:

90% of marketing teams use AI agents in some capacity. Only 23% have them in full production.

So most teams are technically “using AI” the same way they were technically “using social” in 2011. Lots of scattered experiments, no real owner, no production systems anyone can rely on. The 67-point gap between adoption and operationalization is the most useful career signal I’ve seen in years.

The person who closes that gap inside a marketing team is doing something different enough that it’s starting to get its own title. Most people are calling it the Marketing Engineer.

The market is moving fast

Profound made the role official in April. They named Nick Lafferty their Founding Marketing Engineer, though he’d been there for about a year already as their first marketer. The title grew out of the work he was actually doing, not the other way around. Before Profound, Lafferty led growth at Loom and Mailgun (which exited for a combined $3 billion) and ran a $30k/month solo consulting practice.

What he builds at Profound is a decent preview of where this role goes. There are agents that watch media coverage and draft journalist outreach before the PR team has even seen the story. Agents that track competitor moves and update battlecards on the fly. Agents that turn webinars into articles ready for review by the end of the week.

Profound raised a $96M Series C at a $1B valuation in February and is now running a course on Marketing Engineering. They’re clearly betting the title becomes standard within the next few years.

This isn’t speculation. Two years ago Clay coined “GTM Engineer” for the sales-side equivalent of the role. By 2025, GTM Engineer job postings had grown 205% YoY, with companies like Cursor, Webflow, Notion, and Ramp hiring for it at median salaries around $160k. The same trajectory is starting to play out on the marketing side, just a couple of years behind.

State of Brand ran a piece yesterday arguing that the Marketing Engineer title only really works if companies redesign the org chart around it. They’re not wrong. But waiting for executives to redesign org charts has historically been a slow way to build a career. The people developing this skillset right now will be the ones those orgs hire whenever they finally catch up.

What this means for brands

If you run a small or mid-sized brand, this is the most significant shift in marketing economics in over a decade.

Old math: a marketing operation needed a paid lead, an SEO lead, a content lead, a lifecycle lead, and a generalist to stitch it all together. Five hires, $600k+ a year, and 6-9 months before it starts to compound.

New math: one Marketing Engineer with taste and good tooling can stand up most of that operating system in a few weeks. Research, content, paid, lifecycle, reporting. Built once as systems and run mostly on autopilot from there.

This doesn’t mean specialists go away. It means they get added later, where deeper expertise actually moves the needle, instead of being the default first hire. The backbone of a modern marketing function is now somebody who can build.

For founders, this is an edge that wasn’t on the menu 18 months ago. CMOs get something close to a cheat code on their hiring plan. Agencies have a harder question to think about: are they selling output, or are they selling humans? Brands that figure that out before their competitors do are going to compound while everyone else is still posting the old roles.

The arbitrage window

The value of a generalist was never that they could personally do every job. It was that they had enough exposure across the company to connect ideas across functions, plus the taste to know which connections actually mattered.

For 15 years that combination got undervalued because it didn’t scale. You couldn’t personally write every piece of copy, run every campaign, and build every tool. There was always a ceiling on how much one person could ship.

The three things AI tends to handle really well are pattern matching, volume, and baseline execution. Those are also the three things that used to justify hiring more specialists in the first place. Take a generalist with good instincts, give them AI to handle the execution at scale, and you have something that didn’t really exist as a role before.

Most companies are still hiring like it’s 2022 because the org chart hasn’t been updated. Teams that have figured out the new model are quietly outpacing them. There’s an arbitrage window open here, and it’s open until the org charts catch up. Probably another 2-3 years.

How to actually start

If you haven’t tried Claude Code or Codex yet, install one (or both) this week. There’s no shortage of walkthroughs on YouTube to get past the initial setup. I recommend Claude Code but both are proving to be good options.

Pick the most repetitive thing you do in a normal week, the thing you quietly dread on Mondays, and try to automate it end-to-end. Don’t try to make a general-purpose tool that solves the whole category. Just solve your one specific version of the problem. Next week, do another one. Stack the wins.

Save GitHub repos that look useful. Steal from people who are further along than you. When you get stuck, ask Claude to help you debug. It won’t get annoyed, and it won’t tell anyone you didn’t know what you were doing.

The point is to remove manual work from your week. The tools to do it weren’t around 18 months ago, and you don’t need an engineering background to use them.

Final thoughts

12 years as a generalist gave me reps across strategy, content, measurement, and every channel I needed. It was the right career for me. It was also frustrating, because I could usually see what I wanted to build, but rarely had the technical knowledge or time to actually build it.

That frustration is mostly gone now, and what replaced it is the clearest career opportunity I’ve seen in my working life. If you’re a generalist sitting in a similar spot, don’t wait around for HR to invent your new title (or replace you). Start building. The fastest way I know is the same one I wrote about in stop learning AI alone: build a small inner circle of operators and learn out loud together.

If you run a brand, take a serious look at the hiring plan you wrote a few years ago. The math has changed. The team you actually need probably looks different than the one you were planning to hire. If you want help mapping where AI actually fits, the AI Opportunity Sprint is the four-week version of that conversation.

The next couple of years are going to belong to the people who saw this coming and started early.