Arkanza Marketing — Roles & Operations

Plan

7 menit baca

Plan — Arkanza Marketing Knowledge Base build

The map of the whole build: phases, status, and the rules we build by. Pair with Progress (the running handover log). Update the Status column at the end of each session.

North star

A position-based operating system inside one Obsidian vault: each position can open its hub and see JD → tasks → SOPs → the exact sheet it fills → who it escalates to — all traceable to one source. Create → interconnect → publish → maintain.

Scope guardrail (read before building)

  • Position-based, not person-based. Every JD, breakdown, SOP, worksheet, KPI scorecard, and hub describes the position, usable by whoever holds it. No content keyed to a member.
  • Arkanza-contextual, yes. Real brands, marketplaces, tools, KPIs, and the position escalation chain (Digital Advertiser → Superoperator/Deputy → Marketing Lead) stay in.
  • PIC mapping is operational, not KB. Which accounts a person is assigned lives in the App / report sheet, referenced abstractly as “PIC scope: akun yang ditugaskan.” We do not bake the marketplace→person map into the KB. (Who currently holds each position → Org_Structure_and_RACI.)

Phases

Status: ⬜ not started · 🔄 in progress · ✅ done · ⏸ parked

StatusPhaseCore deliverablesSessionsModelEffort
✅1 · Foundationgit · scaffold 09+10 · _Standards templates (JD, Breakdown, SOP, Worksheet, KPI Scorecard, Role Hub, _Conventions) · methodology reference node · Tools index seed · housekeeping2Opus 4.8High
✅2 · Digital Advertiser✅ 16 SOPs · ✅ worksheet + KPI scorecard · ✅ 3 templates · ✅ _DA_Hub · ✅ Product Master v1 xlsx. (Domain nodes Unit_Economics/Benchmarks deferred to when breakeven-ROAS is resolved, post-project per Anton.)3Opus + SonnetHigh → Med
✅3 · Superoperator✅ JD + breakdown docx (revised, position-based) · ✅ 7 SOPs (+ DA-SOP reuse for internal ads) · ✅ worksheet + scorecard (v1 weights ⚠ pending Anton) · ✅ _Superoperator_Hub1Opus 4.8High
✅4 · Marketplace & Affiliate + KOL✅ JD + breakdown docx (position-based) · ✅ 6 SOPs + 2 cross-role playbooks (05) · ✅ built 11_Partnerships/ · ✅ worksheet + scorecard (v1 ⚠ pending Anton) · ✅ _MS_KOL_Hub · ✅ affiliate/KOL tool stubs (Drive home flagged)1Opus 4.8High
✅5 · AI Content Creator✅ JD + breakdown docx (position-based, single-holder) · ✅ 7 SOPs · ✅ Prompt_Library (role-KB node) + QC Checklist template · ✅ worksheet + scorecard (v1 ⚠ pending Anton) · ✅ _AICC_Hub · ✅ AI_Content_Production_Tools stub (products flagged)1Opus 4.8High
✅6 · Marketing Lead✅ JD revised (atasan corrected) · ✅ breakdown (pre-built) · ✅ 11 strategic SOPs (oversight → route down) · ✅ worksheet + scorecard (v1 rollup ⚠ pending Anton) · ✅ _Marketing_Lead_Hub · ✅ 3 domain nodes (Campaign_Calendar, Budget_and_Allocation, Pricing_Architecture) · closes the RACI loop2OpusHigh
✅7 · Calibration & Decisions Lock (final version — Anton)✅ KPI weights + penalty + rollup + cadence/A-B LOCKED & homed · ✅ budget (uncapped ads, margin floor 5%) · ✅ pricing (selling-price bands; COGS→Profit Engine 2026) · ✅ brand portfolio corrected + 4 brand guides filled + positioning ladder · ✅ per-brand affiliate/KOL matrix + KOL fee case-by-case · ✅ Super⇄ML backup · ✅ App discovered → three-layer architecture + zero-cost publish stack decided. Residual data tasks → Phase 8.2OpusHigh
🔄8 · Publish + Govern✅ residual content (ROAS gross-up per marketplace → Unit_Economics/Benchmarks · discount ceilings · sub-brand voice confirmed · handles · KPI targets kept) · ✅ _Hub_Integration_Mapping (KB→Hub seed spec) + KB↔App deep-links · ✅ Quartz+Cloudflare config + steps (setup; Anton connects) · ✅ link/orphan audit + review cadence (_Roles_MOC). Remaining: generate Drive docx/xlsx deliverables + adopt v1 sheets · Anton connects Cloudflare · app-repo agent seeds Hub DB2–3Sonnet + OpusMed

~17–19 sessions total (added Phase 7 · Calibration). Domain-knowledge fill (01–08) is absorbed into phases 2–6 just-in-time — build a domain node when a role SOP needs to link it, not as a separate pass.

You are here

Phase 8 · Publish + Govern — IN PROGRESS (2026-07-13). Done: residual content finalized (Profit Engine mining → per-marketplace ROAS ladder in Unit_Economics/Benchmarks; discount ceilings 3-tier in Pricing_Architecture; sub-brand voice/handles confirmed; KPI targets kept), _Hub_Integration_Mapping seed spec + KB↔App deep-links, Quartz+Cloudflare publish config + steps, link/orphan audit (no orphans, no dup shared knowledge) + review cadence in _Roles_MOC. Remaining: generate Drive docx/xlsx deliverables + adopt v1 sheets · Anton connects Cloudflare (free pages.dev) · app-repo agent seeds Hub DB.

Phase 7 · Calibration & Decisions Lock — COMPLETE (2026-07-11)

All core deferred decisions are locked and homed: KPI weights (all roles) + penalty scheme + Lead rollup + A/B method + cadences + budget (uncapped ads, margin floor 5%) + pricing (selling-price bands; COGS lives in Profit Engine 2026) + Super⇄ML backup. The brand-knowledge layer was built just-in-time: portfolio = Arkanza (3 series) + Dananta/Loka/Sudjiwo + inactive Oamra/Nusantara, architecture = same-management/separate-budget-per-brand, 4 filled brand guides (02) + positioning ladder + per-brand affiliate/KOL matrix (F thread closed).

Key discovery this phase — the App (Marketing HQ, arkanza-hq.vercel.app): the live operational execution layer of the whole roles system (tasks = the SOPs, per PIC, brand+position filters, /hub mirrors the KB with empty buckets, /performance = routine compliance). This settles a three-layer architecture — KB (author/source-of-truth) · App (daily driver + Hub reading surface, links only) · Sheets (data/calc) — and dissolves the H “sheet-vs-App” question (the App owns the workflow stations). Publish stack decided (zero-cost): Quartz + Cloudflare Pages/Access (read) + Google Drive (fillable).

See Progress top block + the Phase 8 kickoff prompt.

Locked decisions (do not reopen)

  • Position-based scope (this doc) — PIC mapping out of the KB.
  • One vault / overlay; knowledge atomic (the 5 buckets are tag/MOC views, not folders).
  • SOP home: 09/<position>/SOPs/ for position-specific; 05 for cross-role playbooks; unified by #type/sop. Daily Marketplace Ops = cross-role playbook in 05.
  • Partnerships: dedicated 11_Partnerships/ — ✅ built Phase 4 (_MOC · Partnerships_Approach · Affiliate_KOL_Criteria_Matrix; KOL_Roster/Affiliate_Programs deferred as placeholders).
  • KPI scoring: direct weighting + achievement capped 120% + rating bands. v1 DA weights 35/25/20/20. Baked in.
  • Guidebook (Fajar Nur’aini, Panduan Menyusun SOP & KPI, 2019) adapted as a methodology reference node; SOP + KPI templates built to its method.
  • Build tracker lives in _Roles_MOC (not the charter).
  • Org/RACI settled; the 5 DA/MS overlaps settled.

Parked / pending (flag, don’t invent)

Most deferred decisions below now converge on Phase 7 · Calibration & Decisions Lock — the “final version” pass where Anton locks KPI weights/targets, per-brand affiliate/KOL criteria, KOL fee/barter policy, breakeven ROAS, A/B method, Superop A-row backup, and log homes together.

  • MS+KOL (Phase 4): KPI weights 30/25/25/20 + numeric targets + penalty scheme; per-brand fit criteria + KOL fee/barter policy → Phase 7. Cadence ✅ confirmed. Affiliate/KOL sheets → ideal v1 built (Partnerships_Tracker_v1.xlsx), awaiting Anton’s adoption to Drive.
  • Superoperator (Phase 3): KPI weights + penalty scheme · Bagian-2 (internal-ads) validation · account/saldo log homes · A-row backup → Phase 7 (except role-holder validation of Bagian 2).
  • AI Content Creator (Phase 5): KPI weights 35/30/25/10 + numeric targets + penalty scheme + role formalization (roster: no prior official JD) → Phase 7. Also: cadence of Laporan Output (daily vs weekly) + Riset Tren frequency (SOP_Laporan_Output_Konten = draft until set); content-output log sheet doesn’t exist (proposed station — KB can build ideal v1 if Anton wants); AI tool products unconfirmed (AI_Content_Production_Tools); AICC backup (single point of failure). “Creative win-rate” KPI candidate needs attribution method + DA↔AICC boundary.
  • Marketing Lead (Phase 6): KPI weights v1 rollup 30/20/20/20/10 + numeric targets + rollup method + penalty scheme → Phase 7 (Lead calibrated last — draws realisasi from the team scorecards). Also: home of master task list / creative log / campaign pipeline (sheet vs App) unconfirmed; nominal budget envelope (Budget_and_Allocation) + price tiers/margin-floor/ discount-ceiling (Pricing_Architecture) pending Anton + Senior Leadership; Campaign_Calendar dates to fill; backup for the Lead A-rows (single point of failure, with Superop) → Phase 7.
  • Breakeven ROAS — ✅ decided 2026-07-10. TikTok = ROAS Target Dashboard (per marketplace); non-TikTok = True ROAS 5.0 (Anton). Remaining: gross-up non-TikTok 5.0 → dashboard target (needs per-marketplace PPN/fee) + home the numbers in 03/Unit_Economics+07/Benchmarks — DA phase.
  • A/B testing method & cadence — Anton + Superoperator decision.
  • Product Master sheet + team product tracker — new builds, don’t yet exist.
  • 3 “dalam pengembangan” items (daily data extraction, ads-log SOP, weekly campaign optim.) — ✅ 2026-07-10: all confirmed to build (Anton co-develops). SOPs authored in DA phase, not gated.
  • Brand slugs (6 primary + 3 startup) — extract when filling Brand_Architecture.
  • KB hygiene (KPI 55→50, 05_MOC, tag namespaces) — ✅ done Foundation 1A.

_Roles_System_Charter · KB_Coverage_and_Gaps · KPI_Definitions · Progress

Tampilan Grafik

Tautan Balik