Data Graveyard

A data graveyard is any dashboard that prioritizes storing numbers over driving decisions. Twelve widgets sit in a perfect grid. Eight color palettes compete for attention. Four time ranges wait for the user to choose. The team calls it data driven and ships it anyway. The user opens it once, feels vaguely informed, and never returns.

The term exists because most product teams treat dashboards as storage instead of tools. They ask what metrics should we show instead of what question does the user need answered. The result is a screen that looks impressive in screenshots but collapses under real use. Attention given to anything becomes attention stolen from everything that matters.

Teams often confuse data graveyards with comprehensive dashboards. They believe showing every metric demonstrates thoroughness. It does not. A comprehensive dashboard that fails the what should I do next test is still a graveyard. The difference is not the number of widgets. The difference is whether the layout contains an opinion about priority.

Salesforce built the archetype. A 4x3 grid of widgets exists because someone asked for each one once and nobody deleted any. Linear's old interface suffered the same fate before the team rewrote it around priority and SLA. Most analytics tools still ship this pattern today. The numbers change but the user behavior does not.

Vercel avoided the graveyard by showing only the deploy timeline and URL on a project view. Stripe surfaces revenue today plus three urgent actions and hides the rest. Cursor keeps its project pane to the file tree, last prompt, and editor with zero telemetry charts. Each example proves the same lesson. Less visible data creates more usable interfaces.

Deploy a data graveyard when your stakeholders cannot agree on priorities and you want to avoid the argument. Watch retention drop. Use it never if you expect users to return to the screen daily. The tradeoff is loud. Some executives will miss their favorite chart. Some power users will request customization. Both groups will thank you later when they actually use the product.

The fix is brutal but simple. Write the one question the screen exists to answer. Delete or hide everything that does not serve it. Rank the remaining elements by how directly they influence the next action. Test it with new users who have never seen the interface before.

Every graveyard shares the same smell. It requires a tour. It ships with an empty state that asks the user to configure widgets. It contains tabs that hide the actual important data on tab three. These are not features. They are symptoms of a team that refused to pick a side.

The data graveyard dies when the team ships the opinion in the default layout instead of the settings panel. Users notice immediately. They return to the screen because it tells them what to do instead of showing them what exists.

Stop arranging widgets. Start answering questions. The dashboard that wins is the one with the strongest opinion about what matters right now.