Marcus was a senior engineer at a 90-person remote-first startup. By any standard metric, his meeting load was reasonable: 3 to 4 hours per day. His colleagues at previous companies had double that. He shouldn't have been struggling.
But he was. Features took longer than estimated. He felt scattered. He described his afternoons as 'a series of half-starts.' The measurement said he had time to think. The experience said otherwise.
What the Calendar Data Revealed
He connected to Tempo and tagged everything: Deep Work / Meetings / Admin. The 3.5 daily hours of meetings were real. But when he looked at the distribution, the problem became obvious immediately.
Those 3.5 hours of meetings were spread across five separate slots throughout the day — a standup at 9am, a sync at 11am, a product review at 1pm, a 1:1 at 3pm, and a team retrospective at 4:30pm on Fridays. The gaps between them: 45 minutes, 90 minutes, 90 minutes, 75 minutes. Four available windows, none longer than 90 minutes.
Deep work requires a minimum of 2 uninterrupted hours to reach flow state. Not 90 minutes. Not 'almost 2 hours.' Two hours, with no anticipation of interruption. Marcus had zero such blocks in his average week.
The Structural Change
Marcus brought the data to his manager. Not a complaint — a proposal. He asked to batch all recurring meetings into the 9am-12pm window, leaving afternoons as a protected deep work zone. His manager agreed immediately. ('I wish every engineer made requests with data like this,' she told him.)
It took two weeks to fully migrate the meeting schedule. Not all meetings were his to move — some required coordination with product and design. But by week three, the pattern held: mornings were meetings, afternoons were his.
The Results
- →Contiguous deep work time per day: 0.8 hours before → 4.2 hours after
- →Feature velocity: subjectively doubled within one month
- →Estimated time to complete current sprint: dropped from 2.5 weeks to 1.5 weeks
- →His own description: "I finally have uninterrupted thought again"
The total meeting hours didn't change. The engineering output roughly doubled. The only variable was contiguous availability. Fragmented time and blocked time are not the same resource — even when the totals are identical.
What This Generalizes To
This pattern shows up across roles — engineers, writers, analysts, designers, anyone whose best work requires sustained unbroken attention. The meeting count isn't the metric that matters most. The length of your longest available block is.