Intern
What you actually do
Pass/fail the spec. Collect artifacts vs PDFs. You are evidence, not a second account manager.
Library / Compare / Fractional ownership vs an agency retainer: who installs the spec
Agencies sell hours, campaigns, and activity. Fractional sells a sequenced leak, weekly ship in the system, and handoff. Both can be honest. Both can be slides. The leak is the same: busy Slack, no pause, numbers untrusted, nothing that runs when the vendor is quiet. Choose by whether you need capacity on a spec you already have, or an owner to install the spec. This is not 'agencies are bad'.
Agencies sell hours, campaigns, and activity. Fractional sells a sequenced leak, weekly ship in the system, and handoff. Both can be honest. Both can be slides. The leak is the same: busy Slack, no pause, numbers untrusted, nothing that runs when the vendor is quiet. Choose by whether you need capacity on a spec you already have, or an owner to install the spec. This is not 'agencies are bad'.
Fractional is right when the leak is unowned, numbers are a vibe, and you need one sequence for 90 days. It is wrong as a cheaper agency that still does random campaigns without pauses.
Agency is right when the spec is written, writers work, and you need production (pages, ads ops, design) against a scoreboard you believe. It is wrong as a substitute for a sale map — then you rented activity on a leak.
Shared leak: retainers without kill, decks as deliverables, Slack archaeology, no system artifact.
How to choose: can a stranger audit a contact and name the Monday number? If no, you do not need more capacity. You need ownership of the model. If yes, buy capacity with the spec as the SOW.
Wiring implication: SOW is tests (write-in, flow, rail), not hours. Agency tickets reference the spec. Fractional week one is instrument if needed.
Intern
Pass/fail the spec. Collect artifacts vs PDFs. You are evidence, not a second account manager.
Operator
Ownership vs capacity. Bind tickets. Protect Monday. Handoff.
CEO
You are buying either an installed layer or hours. If you cannot tell which, you are buying fog. Spec pass/fail is the decision.
Spec existence, leak, pauses, weekly artifact, scoreboard, kill date, hours vs ownership, handoff, Slack as hallway.
If no spec → fractional 90 (or internal owner) before more agency capacity
If spec exists → agency capacity can ship against it; still need an owner of Monday numbers
Either → pause list and kill criteria or you bought anxiety
Handoff notes required if the human is not permanent
Sale map, writers, one number. Pass/fail.
Why: This binary chooses the motion.
Done when: Fail → do not add agency scope. Pass → agency may produce.
Three pauses. Date when this motion stops if X is untrue.
Why: Without these, both vendors become fog.
Done when: Visible to the CEO.
In GHL/Klaviyo/search, not a PDF. Same for agency sprints.
Why: Hours hide in decks.
Done when: Example artifact from week 1.
No extra landing without write-in. No extra workflow without a job.
Why: Unbound tickets recreate the pile.
Done when: A rejected ticket that failed the spec.
Who runs Monday after. Test scripts.
Why: Vendor quiet should not be an outage.
Done when: A stranger script.
No spec, untrusted numbers, too many projects, need a 90-day owner. Wiring: one leak, instrument, pauses, weekly system ship, day 90 booked.
Spec exists, you need production capacity, you will reject off-spec tickets. Wiring: SOW = tests, Monday number stays yours, handoff if they pause.
Retainer fog. Activity without write-in. Slides. Slack as CRM. No kill. The company cannot run Monday without the vendor.
Spec pass/fail. Fail: ownership motion. Pass: capacity on the spec. Both require pauses, artifacts, handoff. Do not buy hours to avoid the page.
If they accept one leak, pauses, artifacts, kill, and handoff — they are operating that layer. The brand on the invoice matters less than the objects.
Strategy without install is a deck. Fractional that does not ship in the system is an agency with better vocabulary. Demand artifacts.
Two is fine if one spec and one scoreboard. Two strategies is the leak.
This week
This week: spec pass/fail, three pauses, weekly artifact definition, kill date. Do not expand a retainer that cannot point at the spec.
Both can be a filing cabinet. Neither is a system until write-in and stages are jobs.
Read ↗Pretty vs native does not matter if two links both take Discovery.
Read ↗DTC lifecycle and coaching spines are different jobs. Two senders is a tax.
Read ↗Rebuilding from memory every client is how you never productize. Blasting a zip is how you never train.
Read ↗Working notes
Pick the one sequence that deserves the next 90 days. Request the file. It arrives by email — not a public dump, not a drip of slogans.
Email and WhatsApp stay open. A call is for installing the system — not for a tour of tools.