Dashboards
Building admin dashboards that ops teams actually use
An admin dashboard succeeds when the ops team stops opening the spreadsheet, the Slack message, and the third-party tool they used to use — and starts opening this thing instead. Most internal dashboards never reach that bar. They’re technically functional and culturally ignored. The difference is rarely about features.
Information density without overwhelm
The temptation when designing an internal tool is to make it “clean” — lots of whitespace, big buttons, marketing-page polish. That’s wrong for the audience. Ops teams come to a dashboard with a specific question they need answered fast. Whitespace puts a cost between them and the answer. Tight, dense, table-heavy layouts win. The goal is “I can answer my question without scrolling” — not “this looks like a product page.”
That doesn’t mean ugly. It means designing for the experienced user, not the first-time visitor. Default to showing the data, not a marketing illustration. A power user spending eight hours a day in the tool will thank you for the density.
Defaults that respect the workflow
The most important UX decision in an admin tool is what the user sees on landing. If the answer is “an empty filter and a Search button,” you’ve already failed — most admin tasks start from a recent context. Show the last 7 days by default, the user’s saved filters, the rows that need attention. If the user has to configure the view every time, they’ll switch to the spreadsheet.
A pattern that works:
- Every view remembers the last filter set, per user.
- Reset is one click.
- New tab opens with the same filters as the current one.
These sound trivial; they’re the difference between a tool people love and one they tolerate.
The underrated power of empty states
Empty states get treated as polish. They’re not — they’re the first impression every new user has. “No results found” is a wasted opportunity. “No tickets match your filter — try clearing the date range, or open the last 30 days” is a hand on the shoulder. The first version of every empty state should answer one question: what should the user do next, in one click?
Internal tools have low budgets and low visibility, but they shape how an entire team works for years. Spend an extra week on the dashboard the ops team uses every day, and you compound the value forever. If you’re scoping an admin or BI dashboard build, this is the lens I’d recommend bringing — the ops team should review it before it ships, not after.
READ NEXT
Why your landing page converts at 0.8% (and how to fix it)
Five concrete patterns that consistently move the needle — and one that almost everyone ships but rarely tests.
From MVP to scale: a senior dev's playbook for SaaS in 2026
The architecture decisions that quietly compound. Why your auth, billing, multi-tenancy, and queue strategy at week 4 will define what week 104 feels like — and the small set of choices that consistently age well.
Cutting P95 latency by 60% on a Next.js + Postgres stack
A profiling playbook: which spans actually matter, where the wins hide, and how to measure without lying to yourself.