Single-Screen Today View

Single-Screen Today View is the home screen that fits in one view and answers what should I look at right now. It uses everything the product already knows about the user, the date, and the business to curate a tight set of signals and actions. Linear built the canonical example. The app opens to your assigned work at the top, an AI written paragraph on project status in the middle, and the next milestone or blocker at the bottom. No charts. No grid of KPI cards. No sortable table. The surface commits to a point of view about what matters most and updates that point of view based on how you actually use the product. Every element on the screen passes the same test. Does this help the user in the next thirty seconds. If the answer is no the element dies. This pattern turns the act of opening the product from a guessing game into a directed handoff. It respects the user's attention budget and spends that budget on forward motion instead of analysis. The product math is brutal. A user who sees their next move in four seconds opens the app ten times a day. A user who faces a wall of charts scans, learns nothing, and bounces.

It is not a condensed version of the old dashboard. It is not a row of four metrics with smaller fonts. It is not a widget dashboard that users customize over time until it looks like the cockpit of a 747. The single-screen today view does not give equal weight to every data point. It ranks and hides. It chooses sides. Where the classic dashboard says here is all the data good luck the today view says here is what you need to handle first. It accepts the trade off that some users on some days will need to click deeper for the full picture. That trade off is the entire bet. The pattern fails if it tries to please everyone by adding back the elements it removed. The discipline required to keep the screen sparse is what separates the teams that ship this pattern from the teams that talk about it in pitch decks but never delete their chart grid. It refuses configuration options that would let the old noise creep back in.

Concrete examples start with Linear in 2024. Their home view shows a maximum of five issues with clear next actions attached to each. Avatars show who else is waiting on you. A generated summary condenses twenty subtasks into three sentences that read like they came from a human coworker. One section highlights anything overdue in red with a single click to resolve or reassign. The whole page loads in under a second and orients the user without any scrolling on standard displays. Cron applies the pattern to scheduling. Open the app and instead of a dense month view with embedded analytics you see your day segmented into focus time, meetings, and breaks. Each segment includes the relevant context pulled from other tools so you do not context switch. Notion Calendar took the idea further by tying database entries and page content directly to time blocks so a marketing team sees campaign tasks next to their standup meeting. Vercel follows the same logic for engineers with a today view that surfaces deployments needing review, live error rates tied to specific services, and flagged logs all in context. Pulse for revenue replaces the finance dashboard with a single screen of today's cash movements, top three deals closing this week, and one anomaly explained in plain text. Each tool demonstrates that when data appears at the moment of decision it drives faster and better outcomes than any separate analytics section ever could.

Use this pattern when your users are operators who check in frequently and need instant orientation. Daily usage by product teams, support groups, and small business owners rewards the today view because it creates a habit loop of open, act, close. The article The End of the Dashboard shows how this beats the old model on the only metric that matters. Time to insight. Ship it as the default first screen. Remove the old dashboard entirely so there is no escape hatch. The subtraction creates clarity. Refer back to the comparison table in the original paper. Primary input for this pattern is identity and time of day. Primary output is one focused answer. It is best for daily operators. The failure mode is hidden depth for power users or analysts who may need links to deeper exploration. Do not use the single-screen today view when the job demands scanning multiple live signals at once like in a network operations center or when the user is an analyst who needs to explore data without guidance. Those cases still belong to the operational status view or the deep analytics workbench. The pattern also breaks if the curation logic stays too naive. Teams must invest in the model that chooses what to show or the screen becomes another thing users learn to ignore.

The single-screen today view forces your team to delete more than you add and that subtraction is what makes the product sticky.