Two jobs are turning into one. A Performance Marketing Engineer spends the budget, builds the systems that measure what it returned, and owns both. No handoff to a separate analytics team, no one else grading their homework after the campaign runs.
For most of the last decade, marketing orgs split spend and measurement into two roles on purpose. A Performance Marketing Manager decided where the budget went. A Marketing Analyst or Analytics Engineer built the dashboards and attribution models that judged whether it worked. Two people, two headcounts, one number they both had a stake in.
That split is disappearing because leaner teams cannot afford to run it as two jobs anymore, and because the tooling that used to require a dedicated analytics hire has gotten a lot easier to run yourself.
Why This Role Exists
The two-job structure has a built-in conflict of interest. The person who wants more budget is rarely the person building the report that decides whether they get it. Splitting spend and measurement across two people does not remove that tension, it just relocates it into a recurring argument about whose numbers are correct.
A Performance Marketing Engineer collapses that split into one person. They are close enough to the spend to know exactly what changed in a campaign, and technical enough to build, not just request, the tracking, attribution, and reporting that proves the result, good or bad.
That conflict of interest is real, but it isn't the whole story. At large enough scale, the split exists on purpose, not by accident. Separation of duties, the same principle that keeps the person who approves a payment from being the one who reconciles it, is a deliberate control once a marketing budget is big enough to matter to a board or an auditor. So is data governance: a public company or a regulated industry needs its tracking pipeline reviewed, access-controlled, and retained on a schedule that one marketer's own scripts usually cannot satisfy alone. Below that scale, the same separation is mostly just overhead.
Even that boundary is starting to move. The same speed pressure compressing decision loops everywhere else in a large org is reaching marketing too, and the response looks less like keeping two departments and more like pushing full ownership down into small pods, each running its own budget and its own numbers end to end, with compliance and data governance sitting above them as an audit layer instead of a second operating team. The org chart gets smaller. The oversight does not have to.
The Job, In Practice
What They Own
- Budget and channel strategy. Allocate and reallocate spend across paid channels (Google, Meta, LinkedIn, programmatic, emerging platforms) based on evidence they generated themselves.
- Measurement infrastructure. Build and maintain tracking, tagging, and attribution pipelines (UTMs, conversion APIs, CRM mapping) instead of filing a ticket with a separate ops team.
- Experimentation. Run A/B tests, incrementality tests, and holdouts to separate correlation from causation in their own results.
- Reporting and forecasting. Turn spend and outcome data into one source of truth (dashboards, cohort views), and forecast CAC, ROAS, and payback with their own numbers instead of borrowed ones.
- Tooling and automation. Build scripts, internal tools, and AI-assisted workflows that scale campaign ops and reporting, instead of repeating the same manual pull every week.
Skills Required
- Hands-on paid media execution across at least two or three major channels (Google, Meta, LinkedIn, TikTok, programmatic), not just oversight of an agency running them.
- SQL, well enough to query campaign, product, and CRM data directly instead of waiting on someone else to pull it.
- Server-side tracking and tagging: GTM server-side containers, conversion APIs (Meta CAPI, TikTok Events API), and UTM and CRM mapping discipline.
- Attribution and experiment design: multi-touch or multi-model attribution, incrementality tests, holdout groups, and knowing which one actually answers the question being asked.
- Budget math: CAC, ROAS, LTV, and payback period modeling, built from their own numbers, not a borrowed dashboard.
- Basic scripting or no-code automation (Python, or tools like Zapier and Make) to connect platforms instead of doing the same manual pull every week.
- AI-assisted building, using tools like Cursor or Claude Code to build and maintain their own tracking instead of filing a ticket with an engineering team.
Increasingly Common, Not Yet Universal
- Setting objectives and guardrails for an AI agent that runs parts of a campaign, then auditing what it did.
- AI visibility and GEO monitoring: tracking how a brand shows up in AI-generated answers alongside the traditional paid channels.
The common thread underneath all of it is comfort moving between a spend decision and a SQL query in the same afternoon. That combination, not any single skill on the list, is what the role is actually testing for.
How It Differs From Adjacent Roles
| Role | Owns budget? | Owns measurement? | Builds own tooling? |
|---|---|---|---|
| Performance Marketing Manager | Yes | Usually delegates | Rarely |
| Marketing / Marketing Analytics Engineer | No | Yes | Yes |
| Growth Engineer | Sometimes | Partial, experimentation-focused | Yes |
| Performance Marketing Engineer | Yes | Yes | Yes |
This Is Already Happening
None of these companies use the exact same title, and that's worth being upfront about. Cast AI is hiring a Growth Performance Engineer who owns paid channels, AI visibility, and performance partnerships in one seat, the closest match to what this piece describes. Marketing Engineer, a job board built specifically for this hybrid category, lists similar roles at Figma and Legora paying $127K to $296K and $188K to $231K. Performance Marketing Engineer, the term this piece uses, is not yet the standard label. It's this piece's name for a job that companies are already hiring for under several different, converging titles: Marketing Engineer, AI Marketing Engineer, Growth Performance Engineer. The titles vary. The pattern underneath, spend and measurement owned by the same technical person, does not.
The people filling these roles rarely come from a degree program. They tend to be performance marketers who ran paid media at a startup with no ops team to lean on and built their own attribution because nobody else would, RevOps or MarTech practitioners who got pulled into media buying because they were the only ones on the team who trusted the numbers, or growth marketers with SQL or no-code skills who got tired of waiting on a data team for reporting that should take an afternoon.
Why Now
Marketing teams are leaner than they used to be. Fewer companies can staff a dedicated measurement function next to a dedicated performance function, so the two get folded into one job req instead of two.
Agentic AI is the bigger shift. Entrepreneur's Dennis Sevilla argues campaigns aren't "a static collection of copy decks" anymore. They're "a complex orchestration of agentic workflows," run by agents that pull a signal from one platform, check it against a second, and trigger an action in a third without anyone approving each step. Campaign execution stops being the job. The human keeps the budget, the strategy, and the judgment call on whether the agent's results are actually working, which is exactly the combination this role is built around.
That shift also lowered the floor for building your own stack. AI-assisted coding tools mean a marketer with no engineering background can build and maintain their own tracking and attribution instead of just requesting it from someone else. Purpose-built tools for this layer do the same for the measurement side, without requiring a platform team.
And the role pays like companies know what they are getting. Plaid's AI Marketing Technologist Lead listing on the same job board runs $170K to $223K, comp that reflects one person doing what used to take two.
The Takeaway
If your team is small enough that this is genuinely one job, and someone is already quietly doing both halves of it, the market now has a name for it. Write the job description above instead of two separate ones, and you'll hire the person who actually exists, not the org chart that used to make sense.
Building your own measurement stack instead of waiting on a data team?
- See how Hardal's server-side tracking fits a marketer-owned setup
- Book a demo and bring your own campaign data