2027 Planning Dashboard
| No. | Track | Track Leads | Scope | Key Results |
|---|
Each downstream 2508 client should pay the fully-loaded GSP cost: hardware + human ops + SW depreciation. Drag the slider between bare hardware (¥263K) and fully-loaded cost (¥711K) to visualise the annual subsidy gap.
T1 ¥263K=bare HW · T2 ¥298K=Booking ref · T3 ¥391K=HW+Ops · T4 ¥584K=HW+Ops+SW Dep · T5 ¥711K=Fully Loaded.
| Client ↕ | Nodes ↕ | Fair billing ↕ | Billed today ↕ | Annual subsidy ↑ | Coverage ↕ | Status |
|---|
Track leads and scope are editable by Admin. Click a cell to edit.
| No. | Track Name | Track Leads | Scope | Key Results |
|---|
Custom PRISM floors by track: T1 ≥52 · T2 ≥55 · T3 ≥55 · T4 ≥50 · T5 ≥45 — 93 projects below floor, ~494MM capacity freed. Every 100 active nodes = ¥48M/yr (fully loaded ÷ 9,072 nodes). Kobo JP routing fix = ¥28M visibility with zero commercial negotiation.
Each downstream 2508 client should pay the fully-loaded GSP cost: hardware + human ops + SW depreciation. Drag the slider between bare hardware (¥263K) and fully-loaded cost (¥711K) to visualise the annual subsidy gap.
T1 ¥263K=bare HW · T2 ¥298K=Booking ref · T3 ¥391K=HW+Ops · T4 ¥584K=HW+Ops+SW Dep (÷ 6,122 dedicated, excl. common pool HW) · T5 ¥711K=Fully Loaded (Budget ÷ 6,122 dedicated).
| Client ↕ | Nodes ↕ | Fair billing ↕ | Billed today ↕ | Annual subsidy ↑ | Coverage ↕ | Status |
|---|
| # | Name | Team | Submitted | Survey Link |
|---|
No matching PICs found.
No projects match your filters.
Billing model · Glide path · Project investment impact · Approved headcount & HW
| Client | Nodes | FY26 | FY27 | FY28 | FY29 | T5 Fair | FY26 Cov. |
|---|
Green ≥100% · Blue 85–99% · Amber 70–84% · Red <70%
Confirmed FY26 billing vs T5 fair share. Source: After Alloc MoM & YoY report 20260703.xlsx — all 44 Search Dept SIDs.
| Client | Nodes | T5 Fair | Billed FY26 | Gap | Coverage |
|---|
Ichiba excluded (114%, cross-subsidising dept). Travel & Beauty: Booking vertical, separate billing.
Review all projects one at a time. Bucket as Plan ✅ · Refine ✏️ · Cut ✗. Progress auto-saves.
| Project | Origin | Track | PIC | PRISM | Bucket | Action |
|---|
| Name | Team | Submitted | Survey Link | Actions |
|---|
| Role | Track | Added | Actions |
|---|
Tier 1 = 100% · Tier 2 = 40% · Tier 3 = 25% · Tier 4 = 10% of Business Impact score. Changes take effect on next M-Score refresh.
| Client | Tier | Description | Impact Weight | HW Nodes | HW ¥M/yr |
|---|
Maps portal client names to their Rakuten SIDs. One SID per row — mark one as Primary for default lookups. Used by the Budget AI chatbot and analytics to resolve client names to cost data.
| Client | SID | SID Name | Primary | Notes |
|---|
Flag domain areas with a history of failed attempts. Red (−2 pts) or amber (−1 pt) applied to Risk dimension when a project name/goal matches.
| Domain | Risk Level | Notes |
|---|
All user accounts and their access levels. Passwords are stored in GCP Secret Manager and not shown here.
If a user needs their credentials, click 📧 to email them securely. Note: Okta SSO will replace these credentials once configured.
| Username | Password | Role | Accessible Tabs | Send |
|---|---|---|---|---|
| admin | •••••••• |
Admin | Leadership, Dashboard, Projects, Project PICs, Budget Planning, Strategy Office | |
| track-lead-${n} | •••••••• |
Track Lead ${n} | Leadership, Dashboard, Projects, Project PICs, Budget Planning | |
| budget-planner | •••••••• |
Budget Planner | Leadership (read-only), Dashboard, Projects (read-only), Budget Planning | |
| dept-manager | •••••••• |
Dept Manager | Leadership, Dashboard, Projects (read-only), Budget Planning, Prioritization |
Click Edit on a row to update its price.
| Hardware Spec | Unit Price (JPY) | Last Updated |
|---|
Click Edit on a row to update its price.
| Seniority Level | Unit Price (JPY) | Last Updated |
|---|
Reference code for key system components. For operational and maintenance use.
Standalone Python scoring engine at /Users/rajasekhar.s/m-scoring/.
Scores 2027 planning survey projects 0–100 across five SDM-derived dimensions.
Run via CLI (python3 cli.py) or Flask API (python3 api.py on port 5002).
This scoring logic is also embedded in scoring.js in this project for in-process use.
Run on any machine inside Rakuten's network to bridge Cloud Run (public) with
jira.rakuten-it.com (private). Polls /api/relay/pending-jira
every 10 s, creates Jira Initiatives, then writes the key back via
/api/relay/jira-result.
Business Impact (max 25 pts) is built from three components — metric type, expected magnitude, and ROI intensity — then multiplied by the client tier.
| Success Metric | Points |
|---|---|
| GMS Uplift, Revenue Uplift | 17 |
| CVR Uplift, Conversion, Purchase Rate | 14 |
| Cost Reduction, Faster Experiments / TTM | 11 |
| Any other metric (Resiliency, CTR, AoV etc.) | 5 |
| No metric entered | 0 |
| ⚠ Score halved if goal text does not corroborate the claimed metric | |
| Impact % | Points |
|---|---|
| ≥ 20% | +13 |
| ≥ 10% | +9 |
| ≥ 5% | +6 |
| > 0% | +2 |
| Negative % on revenue metric | 0 + −3 penalty |
ratio = (client actual spend × impact%) ÷ (effort MM × ¥1.55M) — requires client BvA match, pct and effort
| Return / Cost Ratio | Bonus |
|---|---|
| ≥ 10× | +5 |
| ≥ 3× | +3 |
| ≥ 1× | +1 |
| 0.5× – 1× | 0 |
| < 0.5× | −2 |
Keyword overrides: image/visual search → 10% cap · semantic search → 25% cap (lower of tier vs keyword wins)
| Tier | Clients | Multiplier |
|---|---|---|
| T1 | Ichiba, Travel | 100% |
| T2 | Books, Fashion, GORA | 40% |
| T3 | Other Marketplaces, Other Booking Services | 25% |
| T4 | Other Media Services | 10% |
How to use the 2027 Charter Planning Portal. Click any question to expand the answer.