rate-guard · deliverable

02. Rate Guard — Launch-day enablement kit

Audience: Support tier-1/2, Sales CSAs and partner resellers · Artifacts: 02-kb-bundle.html · 03-troubleshoot-flow.html (practice queue + reference tree) · 04-inapp-tour.html (tour + demo rehearsal) · 3d.html + kb-widget.js (3D streams) · shared data tickets-data.js · Case study: index.html

Fictional client: CargoDesk B.V., makers of the freight-booking platform CargoDesk. Scenario: CargoDesk ships Rate Guard, surcharge protection when fuel prices spike while a quote is open. Enablement must reach Support, Sales CSAs and 60+ partner resellers in launch week.

What the kit does

File	What the learner does	How it is checked
`01-needs-analysis.md`	— (design document)	—
`02-kb-bundle.html`	Search five articles and the launch FAQ; zero-result searches are logged in the gap radar (on Enter or after a pause)	self-test: multi-word search, safe highlighting, gap logging
`03-troubleshoot-flow.html`	Work a launch-week queue of nine tickets from three audiences: search the KB first, route through the triage tree, attach what an escalation needs; a reference tree for free exploration	4 points per ticket (KB 1 · route 2 · discipline 1), 36 in total; pass 30, at most one needless escalation, no ticket closed that needed escalating; one critical ticket
`04-inapp-tour.html`	Guided tour of the quote screen, then a rehearsal: arm, send, answer three customer questions	demo check /4; "demo-safe" at 4
`05-adoption-plan.md`	— (launch cadence, proposed KPIs, iterate loop)	—
`3d.html`	The same queue as three streams into one KB reservoir, triage gates and an escalation tower; leaks show per audience	same scoring

The design decisions worth noticing

The gap loop is practised, not described. Tuesday's German partner asks about white-label quotes; no article exists. Logging the gap drafts an article that answers Friday's Dutch partner. Not logging it makes the same question escalate twice.
One critical ticket. Thursday's invoice looks like a "by design" case but is a real fault (armed, unedited, over threshold, no settlement). Closing it bills the customer money they are owed.
Leaks per audience. The debrief splits results by stream, because an aggregate hides the audience the KB is failing.

Honest limitations

CargoDesk, its customers and resellers are fictional; product UI is a hand-built mock; the triage tree is plausible, not a real company's procedure; figures in the needs analysis are illustrative premises. English master; the two partner tickets carry an English gloss. Not piloted.

Part of I Wayan Usada's learning-design portfolio · all clients and scenarios are fictional composites built from realistic domain patterns · drafted with AI assistance and fully rewritten, fact-checked and owned by a human.

Process deliverables

action-map.md — Action map — the behaviors the artifact must produce, traced to the business goal
storyboard.md — Text-based storyboard — revised to match the build
visual-mockups.md — Visual mockups & style guide — mood, type, key screens, component decisions