Case Study — Metric Comparison
blockmarketingSingle customer result shown as a before-and-after table of four metrics, each row carrying the prior value, the current value, and the delta, under a short context line.
Installation
$pnpm dlx shadcn-vue@latest add https://uipkge.dev/r/vue/case-study-metric-comparison.json$npx shadcn-vue@latest add https://uipkge.dev/r/vue/case-study-metric-comparison.json$yarn dlx shadcn-vue@latest add https://uipkge.dev/r/vue/case-study-metric-comparison.json$bunx shadcn-vue@latest add https://uipkge.dev/r/vue/case-study-metric-comparison.jsonNamed registry:
npx shadcn-vue@latest add @uipkge/case-study-metric-comparisonInstalls to:app/components/blocks/Variants
Loading interactive previews…
Loading interactive previews…
Files installed (1)
app/components/blocks/CaseStudyMetricComparison.vue3.3 kB
<script setup lang="ts"> import { ArrowRight, TrendingDown, TrendingUp } from 'lucide-vue-next' import { Badge } from '@/components/ui/badge' import { Button } from '@/components/ui/button' import { Card, CardContent } from '@/components/ui/card' import { Separator } from '@/components/ui/separator' import { Table, TableBody, TableCell, TableHead, TableHeader, TableRow } from '@/components/ui/table' // `better` records which direction is good, so the delta colour never has to be // hard-coded per row — a falling close time is good, falling accuracy is not. const metrics = [ { name: 'Days to close', before: '9.4', after: '4.2', delta: '−55%', better: 'down' }, { name: 'Ad-hoc data requests / month', before: '212', after: '61', delta: '−71%', better: 'down' }, { name: 'Certified metrics published', before: '0', after: '340', delta: '+340', better: 'up' }, { name: 'Forecast accuracy band', before: '±6.1pp', after: '±3.7pp', delta: '−2.4pp', better: 'down' }, ] </script> <template> <section data-slot="case-study-metric-comparison" class="bg-background"> <div class="mx-auto max-w-4xl px-6 py-20 lg:py-28"> <Badge variant="secondary">Northwind Logistics</Badge> <h2 class="mt-4 text-3xl font-semibold tracking-tight text-balance sm:text-4xl"> Two quarters either side of the rollout </h2> <p class="text-muted-foreground mt-3 text-lg"> Their numbers, their definitions, measured the same way before and after. Nothing normalised. </p> <Card class="mt-8"> <CardContent class="p-0"> <Table> <TableHeader> <TableRow> <TableHead>Metric</TableHead> <TableHead class="text-right">Q4 2025</TableHead> <TableHead class="text-right">Q2 2026</TableHead> <TableHead class="text-right">Change</TableHead> </TableRow> </TableHeader> <TableBody> <TableRow v-for="metric in metrics" :key="metric.name"> <TableCell class="font-medium">{{ metric.name }}</TableCell> <TableCell class="text-muted-foreground text-right font-mono text-sm">{{ metric.before }}</TableCell> <TableCell class="text-right font-mono text-sm font-semibold">{{ metric.after }}</TableCell> <TableCell class="text-right"> <span class="text-success inline-flex items-center gap-1.5 font-mono text-sm"> <component :is="metric.better === 'down' ? TrendingDown : TrendingUp" class="size-3.5" aria-hidden="true" /> {{ metric.delta }} </span> </TableCell> </TableRow> </TableBody> </Table> <Separator /> <p class="text-muted-foreground px-6 py-4 text-xs leading-relaxed"> Measured from their own close calendar and ticket queue. The rollout completed between the two quarters; no headcount was added or removed in finance during the period. </p> </CardContent> </Card> <Button class="mt-8"> Read how they did it <ArrowRight class="ml-2 size-4" aria-hidden="true" /> </Button> </div> </section> </template>
Raw manifest:https://uipkge.dev/r/vue/case-study-metric-comparison.json