Catalog
Same mockup as the pattern page, at catalog width.
setup
Agenda / journey
Show the path of the meeting so the room knows when the ask comes.
consulting-exhibit
Harbor steering
Today we lock the wave plan, then the cutover risks.
Working session · 25 min
Section divider
Mark a new beat without restating the whole deck.
executive-sparse
02
Wave plan
Title / hook
Open the room with one reason this meeting exists.
executive-sparse
Northline board · Q3
Harbor is the last mile we still run on clipboards.
Recap / next steps
Close with owners and dates, not a summary paragraph.
consulting-exhibit
Harbor steering · close
Approve Wave 1 Friday; freeze scope before vendors start.
Ops
Name a single cutover owner by Friday.Product
Freeze Wave 1 SKUs by Friday; no net-new docks.Finance
Hold the dual-run budget through 30 Sep.Next steering: 12 Sep
structure
Comparison (2–3 options)
Show two or three options so the recommended choice is obvious.
consulting-exhibit
Harbor cutover
Wave the cutover; a single weekend is cheaper only on paper.
Big bang
Lowest vendor hours. Highest stockout risk. No rollback.Three waves
Dual-run cost for 10 weeks. Rollback per site. Protects peak season.Pilot only
Safe, but Harbor value slips a year. Board already rejected delay.Northline ops model · fictional
2×2 matrix
Place work on two axes so the starting cell is obvious.
consulting-exhibit
Where to start Harbor
Start where volume is high and process is already standard.
Axes: process standard vs outbound volume
Numbered process
Show a loop or sequence the team must learn in order.
consulting-exhibit
Wave 1 playbook
Scan, stage, ship — train that loop before any other Harbor module.
Ops training · 4 loops, not 14 modules
Timeline / roadmap
Place bets on a calendar so sequencing and no-go windows are visible.
consulting-exhibit
Harbor waves
Midwest in Oct, South in Jan, West after peak.
Do not go live in November
architecture
Reference architecture
Group the running pieces so a failure domain is obvious.
cloud-architecture
Harbor in one DC
Keep scan, queue, and ledger in separate boxes so a jam cannot lock receiving.
Generic boxes · no vendor icons
Swimlane process
Show who does what across functions during a timed procedure.
cloud-architecture
Wave 1 cutover hour
Ops freeze inbound; App drains the queue; Warehouse counts then opens.
Owners are lanes, not colors
Sequence / fan-out flow
Show order and branching of a request so people stop arguing from a box diagram.
cloud-architecture
Ship confirm path
One confirm fans out to ledger, carrier, and customer notice — in that order.
Numbers are order, not preference
System context
Show who talks to whom so the trust boundary is visible.
cloud-architecture
Harbor context
Partners never talk to the warehouse DB; they talk to Harbor API.
Trust boundary: API is the only partner entry
data
As-is vs to-be
Contrast current behavior with the operating shift you want.
consulting-exhibit
Harbor operating shift
Stop reconciling clipboards at close; Harbor writes the ledger at scan.
As-is
Paper dock log. Nightly spreadsheet. Disputes hit finance at month end.To-be
Scan is the ledger. Exceptions only. Finance sees the same object Ops sees.If it is not a behavior change, it is not a to-be
Bridge / waterfall
Walk a number from start to end so the biggest hole is obvious.
consulting-exhibit
Same-day fill
Dwell, not scan quality, is the 12-point hole in same-day fill.
Bridge from theoretical fill to trailing fill
Key findings
List the few facts that justify the title’s recommendation.
consulting-exhibit
Why Wave 1 is Midwest
Standard process plus named owners beat raw volume.
Findings are MECE for this decision only
KPI / metric row
Put a few numbers on the board so one of them is clearly the problem.
consulting-exhibit
Harbor Wave 0
Scan rate is fine; dock dwell is the metric that still kills same-day.
Midwest DC · trailing 4 weeks · fictional
Table exhibit (status colors)
Compare entities on the same criteria with status that encodes the verdict.
consulting-exhibit
infographic
Big-number story
Make one number the slide so the implication cannot be missed.
infographic-editorial
Midwest DC
4.1h
Dock dwell, not scanning, is why same-day still misses.
Trailing 4 weeks · Harbor Wave 0
Annotated chart
Mark the point on a simple chart where the story happens.
infographic-editorial
Same-day fill
Fill fell when dwell crossed four hours — not when we changed scanners.
Annotation is the argument; the bars are proof