Somewhere on your plan for this quarter there are items that have not moved since you wrote it. These are usually the same handful.
A landing page for the campaign that starts in three weeks. A second language for the product pages, which nobody has scoped. The report that would tell you which content actually produced pipeline, which needs your CMS and your CRM to agree on what a lead is. And the AI tools you already pay for, whose drafts your team rewrites before anyone sees them.
Every one of those items is waiting on someone technical, and that person is not coming. Your internal developers are on the product, and agencies sell scoped projects, so nobody sits with you week to week.
There is now a name for the person who clears that list. Several names are in circulation, and they share a lot of ground.
What is a marketing engineer?
A marketing engineer owns the connection between your marketing plan and the systems the plan runs on. They wire your CMS, your analytics, your CRM and your AI tools into something that works as one thing, write down what good output looks like so those tools produce it, and ship the pages, reports and campaigns that used to wait for an engineering ticket. Greg Isenberg put the title in front of a large audience at the end of August, in a video about the role.
This is a real new role, and the newness sits in the skill set. It leans further into development than marketing jobs normally do, and what marks a marketing engineer out against an excellent technical marketer is the ability to connect all the dots.
Are marketing engineer, GTM engineer and growth engineer the same job?
They share a large part of the work. All five describe someone wiring a marketing plan to the systems it runs on, and the job description decides the scope: pipeline systems in one, signup experiments in another, the tool stack or the processes between the tools in the next. Isenberg expects the label to move. “The name is probably going to change but the job won’t.”
Marketing engineer
The marketing team’s own engineer. Campaign pages, the integrations between the CMS, the CRM and analytics, and the AI setups the team works in every day. The title is already in the martech role framework that Scott Brinker describes to MarTech, where the people who build custom apps and digital experiences form one group, often holding titles like “web developer” or “marketing engineer”.
GTM engineer
GTM is go-to-market, and the title is written from the sales side. Cargo, a go-to-market software company, calls the GTM engineer “the engineer who owns the systems that produce pipeline, not the activity that consumes it”, and puts the split like this: “Sales teams talk to customers; GTM engineers build the machine that makes those conversations happen.” Lead routing, enrichment and what the CRM does with a form fill sit here.
Growth engineer
Written around experiments. PostHog, which hires for the role, describes growth engineers as focused “entirely on projects dedicated to driving as much growth as possible”, and its list of the work runs from A/B tests on parts of the signup and activation funnel to building landing pages, SEO on the marketing site and changes inside the product. The first three of those show up on a marketing engineer’s week as well.
Marketing technologist
The widest of the five. MarTech defines a marketing technologist as “someone who designs and operates technology solutions to aid marketing”, and the work includes architecting “the overall marketing stack of all marketing technology products used”. Choosing the tools, administering them, connecting them and watching the quality of the data inside them all sit here.
Marketing ops engineer
The title builds on marketing operations, which is the processes and the data between the tools. GitLab’s published marketing operations job description is a good example of that work: processes “designed, documented, understood, and followed in a way that minimizes bad or incomplete data”, funnel reporting from volume through to average value, and understanding “how the different tools, processes, and strategies influence each-other”.
Four things run through all five titles: pages that need building, systems that need to agree with each other, processes that need writing down, and data somebody has to be able to trust. The weight on each is what a job description sets. Brinker’s groups, as MarTech reports them, “are not always completely separate roles”.
We sell this work under the name Fractional Marketing Engineer, and of the five we think marketing technologist or marketing ops engineer comes closest to what it actually is. If you want to see how we package it, that is our fractional marketing engineer service.
How does an AI agent engineer fit into this work?
AI agent engineer is one of the names used on job listings for someone who builds AI agents, connects them to a company’s systems and keeps them running. Here, we use it as a description of the work, not an official title.
In our marketing work, that means agents that help a team draft and publish through its content management system, with a person approving what goes live. This is the AI part of the broader marketing engineer role described above.
What does a marketing engineer actually do in a week?
The clearest way to see it is to take a request you already make and follow it down both paths, once through the queue you have now and once with somebody in this role.
| What you ask for | How it goes today | With a marketing engineer |
|---|---|---|
| A post that ranks for a term you care about | A writer picks the term, drafts, publishes, and nobody knows for months whether it was the right term | Search Console gets checked, opportunities get ranked by buyer intent, what already ranks gets read, and the draft carries your founder’s actual position |
| A landing page for next month’s campaign | A ticket, a queue, a date somebody else sets | The page is assembled from sections that already exist in your CMS, and the work starts the week you ask for it |
| Which content produced pipeline last quarter | Two exports and a spreadsheet somebody rebuilds every time | Your CMS and your CRM agree on what a lead is, and the report runs itself |
| Product pages in a second market | It gets scoped and quoted, then postponed | The content model already holds a second language, so the work is translation and review rather than a rebuild |
He draws the same line on SEO. “The beginner version is asking an AI to write a blog post about a keyword.” The other version checks Google Search Console, pulls keyword data, looks in the CMS for what exists already, and goes on to “rank opportunities by volume and by buyer intent” before a word gets written. The same tool and the same hour of work produce a completely different thing on your desk.
Where the work actually starts
His example is a software company selling to commercial HVAC contractors.
The obvious campaign for a company like that is something about running your HVAC business better. What the marketing engineer does first is work out which pain gets an owner to take a demo, by reading sales calls, support tickets and churn notes. What comes back is specific: the technician finishes a service call, spots a replacement opportunity, “and the follow-up quote just never gets sent”. Isenberg puts the gap between the two angles plainly. “Stop losing replacement revenue after every service call is obviously a much sharper angle than run your HVAC business better.”
Getting from the first angle to the second is reading work: sales calls, support tickets, churn notes, and a person willing to sit with them until the specific sentence falls out. That half of the role belongs to whoever knows the market best, and in most companies that is somebody already on the marketing team.
Why your AI tools produce drafts nobody uses
Where an assistant has read nothing about your company, it writes the average of everything it has ever seen, and a better model produces a better average. What changes the output is written material the tool reads before it starts: how your buyers describe their problem in their own words, what your company actually thinks, what worked last quarter and what did not. Producing that material sits between marketing and IT.
We run our own marketing this way, on ourselves: a repo of context files, one written job spec per agent, and nothing reaches a live page before one of us has read it and approved it.
What changes for you
Concretely, five things land differently.
- The queue stops setting your calendar. Work that needed a developer starts the week you ask for it, because the person doing it is already inside your stack.
- Publishing stops being a two-person job. Your team drives the CMS with the AI tools it already uses, inside a flow where a human still approves every publish.
- Your AI tools read your own material before they draft, because they are wired to the places your company keeps it.
- The measurement question gets answered. Two systems that never agreed now agree, and the report you keep asking for exists.
- You get your own time back. The hours you currently spend chasing technical work through other people’s priorities go back into your week.
Do you need one?
Five questions, with your own plan in front of you.
- Of the things on your plan that have not moved, how many are stuck for technical reasons?
- Can anyone in your team change a page template or add a field to your CMS without opening a ticket?
- Do your CMS and your CRM agree on what counts as a lead?
- Is there a document somewhere describing how your company writes and what your buyers actually say, or does it live in two people’s heads?
- When your team uses an AI tool, does the output ship or get rewritten?
If most of your blocked items are technical, this is the role. If they are budget or positioning, it is not, and hiring for it will not help.
Hiring one, or working with a fractional team
There is nothing about this work that a team cannot do itself. Isenberg’s own advice is aimed at marketers and founders who want to build it, and that advice is sound.
If you would rather hire this person onto your own team, that is a fine call to make. The question worth answering first is whether you have enough of this work to keep a full role busy. Count the technical items on your plan for the next year, and you will know.
The reason to buy it instead is the reason to buy anything fractionally, which is speed and convenience. Hiring means finding one person equally comfortable in a content model and in a positioning argument. Fractional means the setup exists in weeks. Our fractional marketing engineer service is that arrangement for the technical half of the role: a small senior team inside your stack on a monthly basis, on the marketing team’s side, getting the pages, the tracking and the integrations you need live quickly so a campaign keeps its momentum. The customer-evidence and positioning half stays where it already sits, with your team.
If you want to see where the line falls before you decide, the split between what a marketing team can do on its own and what needs engineering work is worth reading first. The documentation half of this work needs no technical help and nobody’s permission, and doing it tells you how big the other half really is.
Two sentences for your CEO, since you will need them: a marketing engineer is the person who connects our marketing plan to the systems it runs on, so the work we cannot get done stops waiting on the product team. What we get back is speed on everything technical, and a marketing stack our own AI tools can actually work in.