Every package in the six-week sprint, broken into the inputs we control rather than the outcomes we hope for. Read at the top of each L10: number, RAG, one line. Anything red or amber goes straight to the IDS list and gets solved in the meeting, not reported on in the next one.
The operating system around the launch plan: where the work is decided, where it is tracked, where it is measured, and where what we learn goes so it is not lost. One loop, run once a week.
Ryan's Road to Launch: two phases, seven packages, a dated run to Dentown on 19–20 September. It sets the destination, not the week.
Road to Launch + 6-week sprint decksEach package becomes items on the Backlog, tagged Experiment, Project or BAU, ICE-scored, assigned to a sprint with an owner and dates. This is the week's actual work.
Notion · Backlog + Sprints90 minutes, same day, same time. Scorecard read takes 5 minutes: number and RAG only. Everything off track becomes an issue.
This dashboardIdentify the real issue, discuss once, solve it with an owner and a date. Solving three properly beats touching ten. Output is a to-do, a new backlog item, or a decision.
Issues list · no home yetAnything that changed what we believe gets written as one atomic claim with its evidence strength and sources. Not a status update, a belief.
Notion · LearningsLearnings are written to be machine-readable, so the Brain can ground answers in what we have actually proven rather than what we assume.
S3 Brain| What | Tool | Where | Status |
|---|---|---|---|
| The launch planPhases, packages, target outcomes, the dated run to Dentown. | Decks | ~/S3/operations/road-to-launch/ |
Not in Notion yet |
Sprint backlogEvery item of work: Experiment, Project or BAU. ICE-scored, owned, dated, EX-n IDs. |
Notion | Backlog · Sprints | Live |
| Weekly scorecardThe input KPIs we control, RAG'd and commented at the top of each L10. | This dashboard | s3-l10-scorecard.pages.dev |
Draft, agree it today |
| Issues list (IDS)Everything off track, plus anything blocking. Worked by priority in the L10. | Undecided | — | Needs a home |
| LearningsWhat we now believe and how confident we are. One row = one claim, rated Proven / Validated / Directional / Anecdotal. | Notion | Learnings | Live |
| Performance truthModelled, warehouse-backed numbers. The outcomes the inputs are meant to move. | Evidence | s3-dashboards.pages.dev | WIP |
| The BrainAsks questions across the warehouse, reviews, interviews and strategy docs. Every answer cites its source. | MCP | In Claude | Learnings not wired in yet |
This scorecard only holds things an owner can move this week. CAC, sign-ups and retention live on the team dashboard. If an outcome creeps onto the scorecard, someone is being asked to hit a number they do not control.
A to-do, a backlog item, or a decision. An issue that ends as a discussion comes back next week wearing the same clothes.
"Meta CPMs rose" is a status. "Dentist endorsement beats price framing for first-time buyers, Directional, 2 sources" is a learning. Only the second one is worth anything in six months.
Two gaps are deliberate on this page rather than papered over: the IDS issues list has no home yet, and the Learnings database is built to feed the Brain but is not wired into it. Both are decisions for the session, not oversights.
Every metric's target for every week, and the shape it makes. Edit any cell and it pulls straight through to the weekly scorecard. Amber cells are ones we have changed from the draft; reset a row to put it back. By launch is the closing total for a cumulative metric, and the sum of all seven weeks for a per-week one.
| Input KPI | Owner | Target | Actual | RAG | Commentary this week | Trend |
|---|
Personal and business best from the week. Gets everyone present and off their laptops.
This page, top to bottom. Owner reads number and RAG only. No discussion, no explaining. Anything off track is dropped onto the issues list.
The sprint's dated milestones: dentist kickoff 11 Aug, efficacy readout late Aug, Dentown 19–20 Sept. On track or off track.
Last week's to-dos in Notion. Done or not done. Under 90% done is itself an issue.
Identify the real issue behind the symptom, discuss once, solve with an owner and a date. Work the list by priority, not in order. Solving three properly beats touching ten.
Recap to-dos, decide what goes back to the wider team, rate the meeting out of 10. Anything under 8 gets said out loud.