Skip to main content

Deliver Value

Which production workflows scale with people not machines — and what changes that?

§1

Workflow Inventory

The Monday merchandising report still costs 7-13 hours per week. Stage 1 workflow scales with people not tooling. Rework climbs each quarter as data sources diverge. Seven major workflows extracted in Phase 2 — hop count tells the story.

Wake-up pain: Monday report still 7-13 hours and the number cannot be trusted manually

Business Analysis weekly Monday-number report

ARTIFACT

Owner: Finance + Owner/CFO

Hops: 7·Hours/mo: 7-13 hrs/wk

Catalogue saleId production + price publication

ARTIFACT

Owner: Marketing + Buying

Hops: 7·Hours/mo: 16-31 hrs/cycle

Catalogue price loading + POS sync

ARTIFACT

Owner: Finance + Operations

Hops: 7·Hours/mo: 5-10 hrs/wk

Store-level inventory replenishment (×15 stores)

HYBRID

Owner: Store Managers + Buying

Hops: 4·Hours/mo: 22-37 hrs/day

Buyer admin tax (60% of week)

HYBRID

Owner: Buying team

Hops: 7·Hours/mo: 72-120 hrs/wk total

E-commerce publishing + stock-by-store sync

ARTIFACT

Owner: E-commerce / Marketing

Hops: —·Hours/mo: 10-20 hrs/wk

→ Pattern →
§2

Constraint Classification

Stage 1 = ARTIFACT (process choice, not real volume work). Of 7 workflows:

Classification breakdown

  • 0.5 fully REAL — store-manager judgement on the floor
  • 2 HYBRID — buyer deal-judgement (real) + buyer admin (artifact); catalogue selection (real) + production (artifact)
  • 4.5 fully ARTIFACT — Business Analysis, store replenishment communication, e-commerce publishing, catalogue price loading, FX management

Stage 1 attacks the largest ARTIFACT — Business Analysis at 7-13 hrs/wk reclaimed. This unlocks Stage 2 attack on store-communication (15-22 hrs/day across 15 stores) and Stage 3 attack on buyer admin (60% of week reclaimed at point of decision).

→ Pattern →
§3

Automation Candidates

Current → Target → Tool → Dependency

  • Business Analysis manual 7-13 hrs/wk → Auto Monday-number with anomaly flagging → BI substrate (Metabase / Hex / Power BI Premium) → POS export feasibility confirmed
  • Catalogue production + price-sync chain → Single Marketing-Buying sheet auto-loading Catalogue + POS + e-com in one payload → Integration layer on UC1 substrate → Catalogue Platform API access
  • Store-Manager replenishment phone/email → Unified stock-position view per Store Manager with AI-recommended action → Store-Manager UI on UC1 (Stage 2) → UC1 live + Documentation Sprint
  • FX exposure spreadsheet → FX Tracker with scenario modelling and hedge-cover recommendation → Scenario layer on UC1 (Stage 2) → Demand-forecast layer in UC1
  • Buyer admin tax → AI-assisted Buyer Brief per supplier deal → RAG over supplier history (Stage 3) → Documentation Sprint complete + UC1/4/5 in production
→ Pattern →
§4

Capacity Math

Hours/week reclaimed by stage

  • Finance + Owner Monday number — today 7-13 hrs/wk → Stage 1 Month 4 0-2 hrs/wk → Year 1 Month 12 0-1 hrs/wk
  • Marketing time on catalogue post-mortem — today 4-6 hrs/cycle → Stage 1 1-2 hrs/cycle → Year 1 0.5-1 hrs/cycle
  • Store-Manager replenishment comms — today 1.5-2.5 hrs/day → Stage 1 unchanged → Stage 2 Month 9 0.5-1 hrs/day
  • Buyer admin tax — today ~24 hrs/wk per buyer → Stage 1 unchanged → Stage 3 Month 12 ~12-16 hrs/wk
  • Aggregate FTE-equivalent reclaimed by Month 12: 2-3 FTE-equivalent of decision-velocity capacity — without hiring, without restructure
→ Pattern →
§5

Done-When (Stage 1)

Stage 1 is done when: the Monday number arrives at 8am Monday for 4 consecutive weeks; the previous catalogue saleId ROI is reviewed by Marketing + Buying jointly within 48 hours of cycle-end; Finance owns the report (Dream team out of operations loop).

The numbers

  • Number: 4 consecutive weeks of on-time Monday number
  • Date window: Month 4 (Week 16 from kickoff)
  • Owner: CFO + Finance Lead
  • The Ops Director can name the day the Stage 1 milestone is hit — by calendar not by inference
  • Schedule: approve the Week 16 milestone review with CFO; pilot the unified Monday report by Week 6 in dual-run mode
→ Pattern →