Companion to the Kit 02 artifacts: launch cadence, proposed KPIs and the iterate loop. Fictional company (CargoDesk B.V.); nothing below has been run or measured.
`` T-12d T-7d T-3d T-0 LAUNCH T+7 T+30 │ │ │ │ │ │ Needs memo KB bundle Demo rehear- All assets live: KPI snapshot Full review approved in the Support sal: CSAs release-notes link, + gap-radar → fix the weakest macro library pass the support macros, review per stream, or stop + practice demo check partner portal audience what does not help queue live ``
| Audience | T-0 action | Signal to watch at day 30 (proposed) |
|---|---|---|
| Support | KB + practice queue linked from the top macro card; queue run once per agent in launch week | Rate Guard ticket share falling; escalations bounced for missing attachments falling |
| Sales CSAs | Demo rehearsal required before the first customer demo | Share of CSA-sent quotes that are armed |
| Partners | Portal article pinned; tour linked in the onboarding email; DE/NL FAQ live | Reseller organisations with at least one armed quote; partner zero-result searches per week |
Targets are deliberately not set here: there is no baseline in this case. The launch owner sets them before T-0, so they are not retrofitted.
Adoption: share of active accounts with an armed quote · CSAs who passed the demo check · partner organisations activated Quality: Rate Guard ticket share and trend · zero-result KB searches per week, split by audience (gap radar) · needless escalations (escalated tickets tier-2 resolves with a KB link) Business: protected-quote volume, settlements honoured (co-owned with Finance)
Weekly: gap radar + ticket tags per audience → draft an article for any question that appears twice (the practice queue's white-label ticket is this loop in miniature) → monthly KB version bump → quarterly review of tour tips where demos stumble.
A 30-day pre-launch window is the comparison. Attribution is directional, not experimental: without a holdout group, a fall in tickets cannot be credited to the kit alone. A staggered partner rollout (half the regions in week 1, half in week 3) would give a cleaner comparison and is recommended.
Prefer the source? Download the raw markdown.