Get a marketing job that pays $145k to $300k.
The title is marketing engineer: a marketer who builds the agents, automations and pages their marketing runs on.
Eight live weeks from 11 January teach what those job ads ask for. You leave with a working system to show them: tools, pages and agents your marketing runs on, with no hand-coding.
Black Friday offer: 25% off, already applied. A year of CXL, January cohort included.
- Learn exactly what the job ads screen for: AI agents, workflows and measurement.
- Show hiring managers the system, not a line on your CV.
- Certified on what you build, not attendance, with the community there when you get stuck.

From CXL, trusted by 90,000+ marketers at companies like these:
New marketing tools have always created new marketing jobs.
Each time, the marketers who learned to build with the new tool first got the title, the budget and the raise.
AI agents are the biggest new tool marketing has had, and the job they create covers the whole stack: any task that runs by hand can now run as a system.
Marketing engineers don’t wait on others.
Today a landing page waits on a designer and a lead-routing fix waits on engineering. A marketing engineer builds both with AI, the same week, on the CRM and GA4.

Built a campaign engine for a two-person business
“That alone used to take me a whole day working in AI-assisted mode (more days without AI), since I had to write all the copy from scratch because I didn’t like what Claude used to give me. This time it took me about an hour.”
“The Campaign engine made me focus on goals that matter (held calls) instead of just replies.”
Companies are hiring marketers who build. They ask to see the system.
The job ads ask for AI agents, automation and measurement, and for proof you shipped them.
Figma’s job ad said it outright: “We care more about what you’ve built than how many years you’ve spent in any one function.”
456
Job ads for this role, found in the last 14 days.
$163k
Median pay in the 117 job ads that state US pay.
*Job ads: CXL’s Marketing Engineer radar, job ads found in the last 14 days, refreshed every two weeks.
Built her Marketing Brain with Sales, to see what happens after a lead comes in
“I had visibility into what happens on the marketing side, but I was missing a lot of the context that happens after a lead comes in… So I used this project as an excuse to finally bring those conversations together.”
“Here’s how I convinced our VP of Sales to work on this with me. I basically told her: ‘This is going to become our single source of truth.’”
Ran the campaign engine on a real GTM campaign at work
“I completed my campaign engine setup for a small c real-world GTM campaign that I was working on for my job this week and presented it to my boss today… and he loved it!”
Build the systems these job ads are hiring for.
Each job ad names the systems the hire will build: agents, automations, landing pages, lead scoring. The eight weeks build the same kinds of systems on your own stack.

Fixed what AI tools got wrong about his brand, then took the check to clients
“After I fixed my own site and directory listings based on what the accuracy check found, accuracy went from 53% to 71% in eight days.”
“I now run the same AI accuracy check on client brands as a diagnostic.”
Every week teaches a skill the job ads screen for.
We built the eight weeks from the skills those job ads ask for most.
You learn each one by building with it, so you can show a hiring manager the system, not a line on your CV.

Ran the Marketing Brain on company HubSpot and Semrush data
“Pulling deals straight from our HubSpot CRM, plus Semrush data on our content, gave the ICP and the messaging real evidence instead of guesses.”
“The ICP built from deal data didn’t match our own brand guidelines. Seeing “who buys now” next to “who we want next” made the gap obvious.”

Applied the reporting agent to a client’s ad campaigns
“I asked Claude to help me apply Nick’s workflow to help me measure a client’s campaign ads. Claude help me to connect my tokens, Facebook Business ID, the ads UTM’s, etc. So at the end I ended up having a nice new reporting ads workflow for my client.”
Eight weeks. Pick the pipeline job that eats your week, and build the system that does it.
Today that job runs by hand, one chat at a time. From week 4 your system does it on its own, on your CRM and GA4, and by week 8 it is the portfolio these job ads hire on.
Black Friday offer: 25% off, already applied. A year of CXL, January cohort included.
Eight-week cohort, starts 11 January
Live sessions run Tuesdays at 3 PM UTC (10 AM ET, 4 PM CET).
Who this is for and what you need
- This is for you if: you have at least 2 years in ops, growth, demand gen or content, and you run campaigns and read your own performance data. No hand-coding.
- You do not need: prior Claude Code or n8n experience. Every build starts from a template.
- A paid Claude account, Claude Code preferred, plus a free Miro account. The reporting agent and AI visibility builds also use n8n.
- Prep before your first session in the app, suited to all levels. It counts toward your average of 1 hour a day. If you get stuck, we help you through it.
- Any other tool is explained live, with alternatives. You are not locked into one stack.
More builds from members
Demos and quotes from the community, build by build.
Certified on your build and its defense in the community, not attendance.
You post your build in the community and defend it there: reviewers ask questions, you answer them. Along the way, your builds get reviewed like this one, by Diane Wiredu.

Rebuilt the studio’s client content workflow on the Marketing Brain
“Across two clients, factual quality has held up, the drafts are now as good as or better than those we produced before building the brain, and the process takes less time.”
“Operationally, we have reduced working files from up to 20 per article to two and removed several automated review stages.”
How the cohort works
Read more about dates and enrollmentShow less
- The cohort starts Monday 11 January 2027 with setup: about 3 hours
- Live sessions run Tuesdays, 19 January to 9 March, at 3 PM UTC (10 AM ET, 4 PM CET)
- Setup comes first, with technical assistance from the CXL team
- If you join after the cohort starts, you can catch up with recordings and join the remaining sessions, or wait until the next cohort starts. We’ll be there to help you catch up
- One group, start to finish. You build on your own company’s data, or on a sample company if you can’t use it
Read more about your weekShow less
- 90 minutes. The live session: a guest builder or a CXL instructor opens that week’s layer
- The rest of your week. Prep, then you build that week’s layer on your own system
- No hand-coding. You describe what you want, AI writes the code and the SQL, you check the result
Read more about how support worksShow less
- The community is always open and we are in it with you, so a stuck build gets unstuck instead of abandoned
- Technical assistance from the CXL team covers setup support and code review
- Guest builders and experts help us design, review and build these systems
- Every session is recorded, so a missed week is a delayed one, not a lost one
A year of CXL for $1,599 $1,199
January cohort included. 25% off, already applied.
- The January cohort and certification
- Every future cohort in your year
- The community
- 80+ on-demand courses and every minidegree
Black Friday offer: 25% off, already applied. A year of CXL, January cohort included.
Is your company paying? Send this to your manager.
Your system runs on the company’s tools and stays with the company.
Copy this to your manager
Subject: CXL Marketing Engineer certification
I would like to take CXL's Marketing Engineer certification. It is an 8-week live cohort starting 11 January 2027 with setup, an average of 1 hour a day: a 90-minute live session each week, plus prep and building. I would build one marketing system on our own CRM, GA4 and ad accounts that runs unattended for 14 days, and be certified on the build and its defense in the community, not attendance. Everything I build stays with the company. It is $1,199 until 10 January, 25% off, and includes a year of cohorts, system updates, the community and 80+ on-demand courses. Details: https://cxl.com/institute/marketing-engineer/
Who leads and reviews the builds
Hesh leads the live sessions. Peep Laja helped design the program and reviews builds. Diane Wiredu, Tyler Durman, Nick Christensen and Ann Smarty designed or review the builds. Guest builders join as each week opens.
FAQ
When does it start, and how much time does it take?
The cohort starts Monday 11 January 2027 with setup. Live sessions run every Tuesday from 19 January to 9 March at 3 PM UTC (10 AM ET, 4 PM CET). Plan on an average of 1 hour a day: a 90-minute live session each week, plus prep and building at your own pace. Setup comes first, in week 0, opening Monday 11 January: about 3 hours on demand, with technical assistance from the CXL team.
Who is it for?
Marketers with at least 2 years in ops, growth, demand gen or content who want to build what marketing engineer roles ask for, in their current job or the next one.
Do I need to code?
No hand-coding. You describe what you want, AI writes the code and the SQL, and you check the result. Checking is a skill the program teaches, from schema-validated output in week 3 to sanity-checking AI-generated numbers in week 6.
Come with a working laptop and admin rights on the tools you want to connect. You build on your own data, over MCP connections, the standard way AI tools connect to other software. You decide what gets connected and what does not.
What do I build?
One system, a layer a week: a context repo, n8n workflows, LLM steps with evals, and a capstone in one of three tracks. Campaigns goes from brief to reviewed campaign. Content goes from brief to checked draft. AI visibility tracks how ChatGPT, Gemini and Perplexity answer about you, and fixes what they get wrong. On top of it you build a deployed app, a reporting layer and personalization for 50+ target accounts.
How do I get certified?
You pass on seven criteria. It runs unattended for 14+ consecutive days, with 90%+ of runs successful. Its output meets the eval threshold you set in week 1. The result is measured honestly against your baseline. The app and the personalization batch both work. An approval step and a cost cap are set. An assessor can run it from your README. And you defend it in the community, answering the reviewers’ questions.
Show measured improvement over your baseline and you earn the credential with distinction. One resubmission is allowed within 30 days.
What is the defense?
You post your build in the community with a short readout of what your system does and what it moved. Reviewers ask questions in the thread and you answer them. It is the closest thing to a hiring manager asking you to walk them through your work.
How do I prove a system like this is worth it?
A marketer on Reddit asked it directly: “revenue metrics? Calls booked?” Week 6 answers it on your own data: baseline against now, tied to pipeline or hours saved, written up by a weekly report agent. That number headlines your certification.
Is a marketing engineer just marketing ops with a new name?
There is overlap, and the best ops people already work this way. The difference is what you can show: a system you built, running on your own customers and channels, that you can demo and fix. That is what marketing engineer job ads screen for, whatever the title on your CV.
Can I use my company’s data?
Yes, that is the default. You build on your own customers and channels, over connections you control. If you cannot use company data, a sample company dataset is ready for you.
I am in the November sprint. Does it count?
Yes. The sprint builds three layers of this system: the Marketing Brain (week 1), the campaign engine (week 4, campaigns track) and the reporting agent (week 6). You start January with them already built.
What does it cost?
$1,199 until 10 January, 25% off the usual $1,599. It includes the January cohort and future cohorts, every update to the systems, a year in the community, and 80+ on-demand courses and every minidegree on the marketing behind the systems.
How is this different from other marketing engineer courses?
Most teach one vendor’s tool and certify attendance. This one is vendor-neutral, 8 weeks deep, and graded on a system that runs in production, defended in the community.
How is this different from asking ChatGPT?
ChatGPT answers questions. Here you build a system: one that runs on your own customers and channels, that an assessor checks, and that you defend in the community before it counts. ChatGPT will help you do a task faster. It will not run your workflow unattended for 14 days, or check that its output can be trusted at scale.
Can my company pay for it?
Yes. Use the email on this page to send your manager the details. Everything you build runs on the company’s tools and stays with the company.
About CXL
What is CXL?
CXL is a B2B marketing training company focused on AI and conversion optimization. Marketers learn by building real work with live guidance from practitioners.
Does CXL offer AI training for B2B marketers?
Yes. AI for B2B marketing is CXL’s main focus: building AI workflows and agents for research, content, campaigns and reporting, plus visibility in AI search.
Where can I learn conversion optimization?
CXL has trained marketers in conversion optimization and experimentation for over a decade, covering customer research, A/B testing, landing pages and testing statistics.
Proof beats the CV.
Certified on what runs, not on attendance.
Black Friday offer: 25% off, already applied. A year of CXL, January cohort included.




