Operating principles
Seven principles. Each one has receipts somewhere on this site.
Build the system the campaign leaves behind.
A campaign ends and everything it taught you walks out with it unless there's a system to catch it. I judge a quarter by what got added to the machine.
Show the source.
Every number I report carries how it was measured, including the ones on this site. An impressive number with a vague method will eventually blow up in a QBR, so I'd rather show the math up front.
The playbook is what scales, so write it down.
Two people run ten or more concurrent lifecycle programs because the standards, checklists, and prompt libraries exist on paper instead of in anyone's head. Ask for headcount after the playbook is written. The ask usually shrinks.
Put the guardrails in before the AI.
The prompt libraries, QA checklists, and governance existed before any AI touched production work. AI speeds up whatever process you give it, good or bad.
Learn enough to ship it yourself.
When design or engineering bandwidth disappeared, I taught myself HTML and CSS and shipped the pages. The same reflex later built a voice agent, an automation stack, and this site.
The sales handoff is a product.
Speed to lead, routing rules, and the one-click outreach library are features of that product, and sellers are its users. If it lives where they already work, they'll use it.
Take it apart before you trust it.
Cars, martech stacks, attribution models, chicken coops. Once I know what's inside, I can fix it when it breaks instead of waiting on someone who can.