Fractional marketing engineer. A senior marketer who builds the systems.

The term comes from Profound, which introduced the role in 2026 and describes it as “a full-stack marketer with the skills of a builder”: someone inside the marketing team who has learned to write code, connect APIs and build systems and agents. Their short version is that marketing engineers “build at the system level to automate at the campaign level.”

The important word is marketer. As Justin Taylor put it in The Landing Pad, knowing what to build is harder than knowing how to build it. A developer can automate a report. A marketing engineer knows which report changes a budget decision, and builds that one.

Not to be confused with two older meanings that share the search results: the engineer who supports product marketing at a manufacturer (most of the Seek listings in Australia), and the academic field of marketing engineering, which is about decision models.

Every new channel adds work and retires none. Social, short video, creators and now AI search all sit on top of Google, email and the website. Profound’s line is that the work grows exponentially while headcount grows linearly (The rise of the marketing engineer).

What changed in 2025 and 2026 is that AI agents can now research, draft, analyse and carry out tasks reliably enough to hand them real work. Profound splits the job in two. First, speed up the playbook you already have: find the steps that repeat, the rule-based decisions and the handoffs, automate them, and keep a person on anything that needs judgement or taste. Second, build things your team could not do before at any headcount.

In a smaller business that second half matters most. A founder-led company will never hire a data engineer, a web developer and an SEO analyst. One person who can build what each of them would have built is a different proposition.

These roles overlap, and in a small company one person often covers two. The difference is what each one builds.

Websites as code. Every site I build now runs on Astro and Cloudflare, with the copy in a GitHub repository. An agent can add a suburb page, fix a headline or ship a new resource in minutes, and every change is a commit someone can review or roll back.

Search and AI visibility. Search Console, DataForSEO and IndexNow read and written through their APIs, so rankings, AI Overview citations and new pages are checked on demand instead of in a monthly export. Structured data, share images and an llms.txt file are generated from the same content, so they never drift out of date.

Ads and creative operations. Reporting that pulls Google and Meta together, a library of saved competitor ads, and creative briefs written from what the account says is working.

Reporting and memory. Dashboards that refresh themselves, and a plain-text knowledge base of the brand, the customer and past experiments that agents read before they write anything.

The outbound side (lists, signals, sequences) is GTM engineering. I do both, and they work better planned together.

Worth asking, since Profound sells AI search software and a new job title creates buyers for it. Its course, certificate and job board are its own products. The name might not last.

The work is real, though. A year ago my websites were edited by hand, my SEO data came out of tool logins and my creative research lived in a paid app. Now one conversation can pull search data, check an ad account, edit a website and ship the change. Whatever the title ends up being, someone in your marketing team needs to be able to do that.

Guardrails matter more than tools. Agents get narrow jobs, every change goes through review, and results are measured on qualified leads and revenue rather than how much got produced.

The first salaried roles are in the US. Profound lists its own marketing engineer at US$135k to US$200k, and Stripe’s comparable marketing role is advertised at US$132k to US$198k. There is no reliable Australian salary data yet, but a senior hire with these skills would sit well into six figures before super and tools.

My retainer is flat, from $1,500 a month, and covers the strategy and the build. Software and API costs are billed at cost and sit in your accounts. The first month is a paid trial: if the first system is not live within 30 days, you can walk away and I refund the retainer.