Six areas I own, roughly in the order a program starts wobbling if nobody's watching them.
Journey Builder, Automation Studio, Data Extensions, Content Builder, AMPscript and SSJS. Everything else depends on this layer. Get the data model and automation logic wrong, and no amount of good copy saves the send.
I build for six months from now, with naming conventions that still make sense later and Data Extensions that don't quietly duplicate records. Automations should fail loudly, so you catch the problem before it costs you a send. I reach for native platform tools first, and I'll tell you when an external script is the better call.
Scoring, grading, engagement programs, and the marketing-to-sales handoff. This is where a lot of marketing ops' real revenue impact lives, and where a shared definition of "ready" usually never gets built.
I've rebuilt scoring models from scratch, sat down with sales to redefine what "MQL" means instead of deciding it alone, and wired automated routing so a lead doesn't sit in a queue because a rule broke three months ago. I think of the handoff as a relationship between two teams, not a data pipe. It works better that way.
List hygiene, suppression logic, bounce and complaint handling. Unglamorous work, and it determines whether your program lands in an inbox at all.
I wrote and rolled out a formal subscriber hygiene SOP covering suppression rules, re-engagement thresholds, audit cadence, and ownership. I treat hygiene as a standing process with a named owner, because that's what keeps it from quietly decaying.
KPI frameworks built around delivery rate, click-through rate, and opt-out rate. Open rate has gotten shaky since Apple Mail Privacy Protection, so I treat it as background noise, not a headline number.
I build reporting around the decision it's supposed to inform, rather than the number that looks best in a screenshot. If a metric can't help someone decide something, it doesn't belong at the top of the report. More on the open rate problem →
Pages treated as a living asset, tested and iterated against real campaign performance.
Landing pages share the same discipline as everything else on this list. If you're not measuring what happens after the click, the email numbers upstream don't tell you much. I hold page performance to the same "what does this number let us decide" standard I hold email reporting to.
Keeping the automation platform, CRM, and everything in between speaking the same language. This layer stays invisible when it works and gets very visible when it doesn't.
Most failure points in a marketing ops stack live at the seams between tools. That's where I spend a disproportionate amount of time on purpose, plus SQL-level data work when the platform's native reporting can't answer the question someone's asking.
Anonymized examples of this work in the field. Client details are generalized. The mechanics are real.
View Case Studies