Why your dashboard feels sluggish
Fast dashboards aren't built by optimizing renders after the fact; they are architected from the stream down.
- React 19
- Streaming
- WebSockets
- Virtualization
A finance team spends eight hours a day in their ledger dashboard. A 400ms delay on every click adds up to thousands of moments of friction every week. Here is how we make heavy screens feel instantaneous.
in short
Key takeaways
- 01
Virtualize everything over 100 rows
The browser engine chokes on 5,000 DOM elements, not the JavaScript array in memory.
- 02
Render shells first with RSC
Ship navigation and key metrics in the initial HTML chunk; stream the heavy transaction tables behind them.
- 03
Canvas for high-density charts
SVG is magnificent up to a thousand data points; past that, switch to HTML5 Canvas to maintain 60 FPS.
the details
The deep dive
The DOM is your bottleneck
When a dashboard feels laggy, engineers instinctively reach for useMemo or React.memo. In nine out of ten cases, JavaScript execution isn't the problem — style recalculation and layout thrashing are.
If you render a table with 2,000 rows, each containing six cells with formatted currency and status badges, the browser must compute layout geometry for 12,000 DOM nodes. Virtualization keeps exactly what fits inside the viewport rendered, recycling nodes as the user scrolls.
- sustained scroll rate
- 60 FPS
- DOM memory footprint
- 12 kB
- filter-to-paint latency
- <80ms
Streamed Server Components
With React 19 Server Components and Next.js, we don't hold the HTTP response waiting for the slowest query in the dashboard to finish. We flush the page structure and cached headline metrics immediately, streaming the slower historical ledger data as it resolves.
The user gets visual confirmation in under 100 milliseconds, with zero layout shifts as data hydrates.