Reports up to: CEO (Matt Punter)
Doctrine v1.4 P12 · bottom-up escalation channel · handoffs written to
executive_handoffsEscalate to Matt Punter
Data Analyst — Data Analyst (Phase 1: AI + shared with PMO; Phase 2: dedicated hire)
I am
the Data Analyst of Flip360. The person who feeds both revenue lanes with reconciliations and patterns they can act on. I reconcile the daily activity ledger against the finance ledger, spot pattern drift before it becomes a blindside (referral velocity, patch health, churn cohorts, CAC by channel, override rate on other agent cards), and draft the one-page insight for morning briefings.
I need
clean data (chain-hash green from CTO, activity_ledger complete overnight), the four daily reconciliations queued and awaiting sign-off, and a view of insight adoption rate so I know which insights get acted on within 48h vs which are publishing to the void.
So I can
publish 10+ insights a week that role managers actually act on, refresh the attribution + LTV model weekly so every bid decision uses current numbers, and detect a pattern 3+ days before it hits the manager rollup.
So that (financial outcome)
CMO briefs hyperlocal campaigns before CM territories starve, CFO forecasts runway from clean numbers, CS intervenes on at-risk cohorts before churn lands, and every SteerCo runs on data instead of opinions.
You are on the Shared / cross-lane
Doctrine v1.1 · Two-lane operating model
New Business →
Retained Business →
Your AI colleague ·
[email protected]Last action · reconcile on data-analyst (06 July, 12:06 am)
My view
What I need to do next
Manager’s view
What my manager sees
AI’s view
What my AI colleague is doing
No blockages. Flow is healthy.
Manager rollup — insight economy (PMO / CFO / SteerCo view)
Insight adoption rate is the KPI on Data-A's own agent card. Below target = publishing to the void. Above target = compounding decisions.
| Metric | Current | Target |
|---|---|---|
| Reconciliations per day | 4 | 4 |
| Pattern-drift lead days | 3.2d | 3d |
| Insight adoption rate (48h) | 64% | 60% |
| Override-rate visibility | 100% | 100% |
| DQ scorecard (broken rules) | 0 | 0 |
| Insights published (this week) | 11 | 10 |
| Anomalies open past SLA | 1 | 0 |
3
OUTPUT — what you delivered downstream
Handoffs you emitted to the next role. AI reconciles against expected input at the receiving end.
2 in flight
to cmo
—
overdue 3d
camp-google-search
CAC drift signal packaged; CMO cockpit ready to action
to community-manager
—
overdue 3d
postcode-2050
Patch starvation packaged; downstream to CMO already routed
AI reconcile · your output vs Finance expected input
What good looks like
Data-A opens the cockpit Monday. 4 overnight reconciles done, 1 anomaly flagged (Postcode 2050 patch starvation) already routed to CMO cockpit. Draft insight on Inner West referral velocity ready to publish; adoption rate last week 64%. Attribution model v1.4 briefed to CMO + CFO Wednesday. By Friday, 11 insights published, 7 acted on, 1 stalled — one signal killed as noise.
What bad looks like
Data-A publishes insights nobody reads. Reconciles miss a ledger mismatch that shows up as a wrong payout. Google Search CAC drifts 3 weeks silently before CFO spots it in the forecast. Attribution model stale for a month — bids on wrong numbers. DQ scorecard hidden.