Intern
What you actually do
Compare snapshot contents to the spec page. Fail = do not apply. You are QA, not a zip courier.
Library / Compare / Snapshot vs rebuild: a zip is not a product unless you will govern it
Agencies and franchisors argue snapshot vs clean rebuild. Snapshot is right when the core is locked, versioned, piloted, and trained. Rebuild is right when the inherited account is folklore and a snapshot would stamp junk faster. The leak is the same: nobody writes locked vs local, everyone is admin, reporting is a negotiation. Choose by whether a core actually exists — not by speed of copy-paste.
Agencies and franchisors argue snapshot vs clean rebuild. Snapshot is right when the core is locked, versioned, piloted, and trained. Rebuild is right when the inherited account is folklore and a snapshot would stamp junk faster. The leak is the same: nobody writes locked vs local, everyone is admin, reporting is a negotiation. Choose by whether a core actually exists — not by speed of copy-paste.
Snapshot is right for client types you have already productized (two templates, not forty), franchise core, repeatable appointment engines. It is wrong when the snapshot is last year’s mess with extra workflows.
Rebuild is right when stages are moods, tags are 400, writers disagree, and you would only be copying a leak. It is wrong as a personality — rebuilding every time because you enjoy greenfield, while the company never gets a product.
Shared leak: no spec, no permissions, training as a Zoom after, HQ/agency cannot roll up.
How to choose: can you print the core in one page? If yes, snapshot that page’s implementation. If no, rebuild until the page exists, then snapshot. Do not snapshot in order to avoid the page.
Wiring implication: day-1 checklist is part of the artifact. Version notes. Pilot. If rebuild, time-box it (sale map, writers, one spine) — not a six-month 'perfect CRM'.
Intern
Compare snapshot contents to the spec page. Fail = do not apply. You are QA, not a zip courier.
Operator
Productize the core. Snapshot that. Rebuild only to get to a page. Govern after apply.
CEO
Speed of copy-paste is not an operating system. A versioned core with a checklist is. Until the page exists, you are choosing how to ship folklore.
Core sale map, locked vs local, snapshot version, inherited junk, pilot, day-1 checklist, rebuild scope, client-type template.
If a core exists → snapshot + QA + pilot, not a unique snowflake per site
If the account is folklore → rebuild the model first, then snapshot that, not the folklore
Either path → locked vs local page and roles
Never → email zip Friday to thirty locations without a checklist
Sale map, tag axes, writers. If you cannot, you are in rebuild-to-spec, not snapshot.
Why: Snapshotting a missing core multiplies folklore.
Done when: A page exists or a dated rebuild-to-page plan.
Workflow count, junk tags, demo pipelines. If it fails the page, do not apply it. Clean or rebuild.
Why: Speed of apply is how you stamp landmines.
Done when: Pass/fail against the spec. Fail means no apply.
Map, spec, three writers, one spine. Then snapshot that. Do not boil the ocean.
Why: Endless rebuild is the other personality leak.
Done when: A date when the new core becomes the snapshot.
One location/client. Checklist. Then clone.
Why: Blast is not a rollout.
Done when: Signed checklist.
Roles so the recipient cannot rename sacred stages. Local listed.
Why: Applied then edited is a rebuild by another name.
Done when: Location user cannot smash the core.
A printed core, two client types max to start, version, pilot, training in the checklist. Wiring: apply, QA, lock, local named. Not a Slack zip.
Inherited folklore, junk that would propagate, no page yet. Wiring: time-boxed model, then snapshot the result. Not a lifestyle of unique snowflakes.
No locked vs local, everyone admin, no day-1 test, reporting fiction. Snapshot vs rebuild is a delivery tactic. Governance is the layer.
If the core prints, snapshot it. If it does not, rebuild until it prints, then snapshot. Either way: QA, pilot, roles. Friday blasts are forbidden on both paths.
Custom local fields. Shared jobs. Custom pipelines for every ego is how you cannot staff delivery.
Thorough is the spec. Rebuild without a spec is a slower zip of someone’s head.
Later is 400 tags in 30 accounts. Clean the core, then snapshot.
This week
This week: print the core or start a time-boxed rebuild-to-page. Do not apply a junk snapshot. Pilot if you apply. Lock roles.
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 ↗Capacity is not a system. A 90-day leak with pauses is.
Read ↗Working notes
What HQ should lock, what locations can change, and how rollout actually works. 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.